1

QuickTime™ and aTIFF (Uncompressed) decompressor

are needed to see this picture.

Technology-driven entrepreneurship: how technology shapes the way entrepreneurs

start and run their businesses

5th Inter-RENT Online Publication

Editor Luca Iandoli

European Council for Small Business and Entrepreneurship (ECSB)

2

© European Council for Small Business and Entrepreneurship (ECSB) and Authors Published by European Council for Small Business and Entrepreneurship Turku, Finland, march 2008 5th Inter-Rent Online Publication

3

TABLE OF CONTENTS Editorial preface: Technology-driven entrepreneurship: how technology shapes the way entrepreneurs start and run their businesses

Luca Iandoli Inward Foreign Direct Investment and ICT: Are they a Joint Technological Driver of Entrepreneurship?

João Leitão, Rui Baptista Creating Technology-based Companies – about the Significance of Effectual Logic in the Early Phase of the Venturing Process

René Mauer, Malte Brettel Funding New Technology-Based Firms: beyond Venture Capital, the Role of Debt

Tommaso Minola, Marco Giorgino Extending the Knowledge Spillover Theory of Entrepreneurship for Analysis of the Australian Biotechnology Sector

Julienne Senyard, David Pickernell, Nick Clifton, Adrian Kay, Robyn Keast

4

TECHNOLOGY-DRIVEN ENTREPRENEURSHIP: HOW TECHNOLOGY SHAPES THE WAY ENTREPRENEURS START AND RUN THEIR BUSINESSES

Started in 2004, the INTERRENT is an on-line ECSB publication initiative aimed at creating a

forum where ECSB members, and particularly post-doc scholars at the early stages of their academic

career, can debate around a selected theme from the previous RENT conference, leading to the

publication of a number of high-quality articles. This initiative has an exciting potential for ECSB as

an organization and, more specifically, for its members. It tries to provide support to young scholar

to improve their work by promoting the progression from conference papers into published articles.

Furthermore it contributes to develop further debate and understanding of a new topic each year.

The fifth edition of the INTERRENT contains a selection of four papers presented at the 2007 RENT

conference in Cardiff (UK) on the following theme “Technology-driven entrepreneurship: how

technology shapes the way entrepreneurs start and run their businesses”.

The idea behind this topic proposal is that technology is not only a driving engine for economic

growth and entrepreneurial ventures but also that it acts like a powerful force able to shape

entrepreneurs’ behavior, values and mindsets. In this perspective technology is not a neutral tool, but

something that at the same time supports and constraints human action, thinking and even deepest

values. Each technological innovation brings with it a wave of new start-ups and new entrepreneurs,

but probably the way entrepreneurship is deployed changes in any major technological revolution.

The literature on technological entrepreneurship is often obviously focused on the most recent

technological trends and high-tech companies, but we think that interesting insights could come from

comparative and longitudinal studies about how entrepreneurship develops in different technological

areas (low and high tech) and in different times.

A few words to illustrate the criteria used for papers selection and the review process. The criteria for

papers selection have been the following: authored by at least one post-doc, quality, coherence with

the topic, and absence of overlapping with the RENT anthology. The papers underwent a two steps

review process: a cross-review phase, in which each participant has reviewed his/her peers’ work

using the on-line forum available on the ECSB web site www.ecsb.org; a second step with a more

traditional blind review by anonymous experts.

Papers include a certain variety in terms of industries like ICT and biotech. All report empirical

studies, with a mix of qualitative and quantitative approach. They deal with several topics like access

to capital for new technology-based companies (Minola and Giorgino), the management of

uncertainty in the venture creation process of technology-based ventures (Mauer and Brettel), the

5

spill-over effects in knowledge-based industries (Senyard, Pickernell, Clifton, Kay and Keast), and

the investments in IT as drivers for entrepreneurship (Leitão and Baptista).

A word of thanks goes to all the participants to the INTERRENT 2008 and the anonymous reviewers

for the quality of their efforts, their responsiveness and the ability to cope with the constraining

deadlines imposed by a tight schedule.

The 2008 INTERRENT Editor

Luca Iandoli1

University of Naples Federico II, Italy

DIEG, Piazzale Tecchio 80, 80125, Naples (Italy)

iandoli@unina.it

1 Currently a Fulbright Research Visitor at the Center for Collective Intelligence, Massachusetts Institute of Technology, Cambridge (MA), USA

6

INWARD FDI AND ICT: ARE THEY A JOINT TECHNOLOGICAL DRIVER OF ENTREPRENEURSHIP?

João Leitão Instituto Superior Técnico

Technical University of Lisbon, Portugal Ph: + 351.214.233.585, Fax: + 351.214.226.901

E-mail: joao.c.leitao@tagus.ist.utl.pt

Rui Baptista Instituto Superior Técnico

Technical University of Lisbon, Portugal Centre for Innovation, Technology and Policy Research

IN+, Lisbon, Portugal Max Planck Institute of Economics, Jena, Germany

Ph: +351.218.417.379, Fax: +351.218.496.156 E-mail: rui.baptista@ist.utl.pt

ABSTRACT

This paper investigates whether inward Foreign Direct Investment (FDI) and investment in Information and Communication Technologies (ICT) advance the development of entrepreneurial activity in the host economy. We propose that the combination of inward FDI with investment in ICT is a joint technological driver of entrepreneurship. Under a feedback causality context, a Co-integrated Vector Autoregressive approach is used to examine the ‘pull’ effect of ICT and the ‘push’ effect of FDI. On the one hand, ICT ‘pulls’ FDI; on the other hand, FDI ‘pushes’ investment in ICT. Under a neo-Schumpeterian approach, the long term economic relationship among entrepreneurial activity, FDI and ICT drives creative destruction through the creation of further SMEs, thus revitalizing the entrepreneurial innovative capacity of the host economies.

7

INTRODUCTION

Inward Foreign Direct Investment (FDI) is considered a strategic action for the advancement of

entrepreneurial activity in domestic firms networked with multinational companies (MNCs). The

spillover effect of MNCs intensifies competition, enhances innovative capabilities, aids the

development of Research and Development (R&D) activities, and also contributes to reducing social

exclusion (Parker, 2005, 2006). The seemingly weak evidence on positive effects of the presence of

MNCs on entrepreneurial activity might hide their significant impact on industrial re-structuring

(Acs & Szerb, 2006; Barbosa & Eiriz, 2007).

Studies of the impact of inward FDI on entrepreneurial activity are relatively scarce. Of

particular importance are those by Sleuwaegen & Dehandsschutter, 1991; Jovanovic, 1994; Van

Praag & Cramer, 2001; De Backer & Sleuwaegen, 2003; Barbosa & Eiriz, 2007; and Burke et al.,

2007. However, these studies focus on the impact of imports and/or inward FDI on entrepreneurial

activity or the survival of start-ups without paying specific attention to the mechanisms through

which such spillover effects occur. This particular study contributes to the literature by focusing on

the role played by investment in information and communication technologies (ICTs) in driving

spillovers originating in FDI.

There is growing empirical evidence of the dynamic role of entrepreneurial activity in promoting

innovation, economic growth and employment (Audretsch et al., 2006; Van Stel, 2006; Fritsch &

Mueller, 2004; 2008). Acs et al. (2006) speak of entrepreneurship as a conduit for spillovers of

knowledge into the economy, enhancing productivity and growth. But entrepreneurial activity itself

is also driven by spillovers, acting either through the transmission and availability of innovative

ideas, or through the existence of an infra-structure which supports new entrepreneurial efforts. We

propose that two dynamic drivers of spillovers that increase entrepreneurial activity are FDI and

investment in ICT.

Schumpeter’s (1940) theory of creative destruction proposes that non-innovative firms and

products are replaced with innovative ones. The concept of innovative firm may here include those

firms who take advantage from opportunities arising spillovers available in the environment. The

environment here comprises both physical infra-structure and the pool of knowledge and demand for

new services generated by existing firms. Investments that contribute to the creation of sources of

spillovers in the environment drive creative destruction whenever new or existing firms take

advantage of these spillovers. In order to take advantage of the aforementioned spillovers, new and

existing firms are required to make more investments that, in turn, generate more sources of

spillovers.

In the present paper, we examine evidence of this sequence of dynamic relationships that are

established between entrepreneurial activity and its dynamic drivers, looking more specifically at

8

FDI and ICTs. The main contribution of the paper to the literature of entrepreneurship and

technological change is the analysis of inter-relationship between these two dynamic drivers of

entrepreneurial activity, and of their joint impact on entrepreneurial activity and, indirectly,

economic performance.

More specifically, we assess the causality relationships that are established between ICT

investment and FDI, and economic variables used in the literature to assess entrepreneurial activity –

the business ownership rate – and economic performance – gross domestic product (GDP) and the

unemployment rate. The present study examines data for two different countries – Finland and

Portugal, providing a comparative analysis between two European countries with different

technological profiles: Finland2 (a high-tech producer) and Portugal3 (a low-tech producer) that

experienced significant FDI and investment in ICTs.

In order to assess causality in the relationships between these variables, a Co-integrated Vector

Autoregressive (CVAR) approach is used. The CVAR approach provides a dynamic analysis of the

experience effects that result from the adoption of public policies oriented to FDI and ICT on the

performance of two European countries with different technological profiles.

The remainder of the paper is divided into five additional sections. In the second section, a brief

literature review motivates the need for linking the FDI ‘push’ and the ICT ‘pull’ effects within the

context of public policies for fostering entrepreneurship, looking at the role played by FDI as an

investment ‘push’ and ICT as a technological driver. The third section examines two streams of

research on spillovers affecting entrepreneurial activity and economic performance: spillovers

originating on FDI; and the catalysing role of enhanced ICT infra-structure in fostering other sources

of spillovers leading to innovation and entrepreneurship. As a product of the analysis of these two

streams, the hypotheses of the study are presented. The fourth section introduces the data used in the

study and describes the methodology. The fifth section presents the empirical findings and discusses

the main results. The final section discusses the results and presents some possible explanations for

the main findings of the study, while referring its limitations.

LITERATURE REVIEW

According to Baumol (1968:p.69), public policies should be directed to “induce the appearance

of increased supplies of entrepreneurial skills” and the policy-maker should be “interested primarily

in what determines the supply of entrepreneurship and in the means that can be used to expand it”.

The theoretical model proposed by Baumol (1990) for the determinants of the ‘allocation’ of

2 In this context, the country has a major manufacturing industry of Electronic and ICTs. 3 In the current framework of analysis, the country presents generally production activities oriented to the use of earlier or less developed technology.

9

entrepreneurship suggests that the regulatory framework plays an important role in the determination

of the success of entrepreneurship, whether it will be a productive or an unproductive driver of the

national productivity growth.

A related question that centres on how public policies vary according to the level of economic

development has been widely explored in the economics literature, under different approaches: (a)

economic development (Lucas, 1993), (b) regional science (Acs & Storey, 2004), and (c)

entrepreneurship (Acs, 2006; Acs & Szerb, 2006). The empirical studies by Van Stel et al. (2005),

Wennekers et al. (2005) and Acs & Varga (2005) revealed that, for highly developed economies,

entrepreneurship has a positive effect on economic growth. Nevertheless, developing economies may

be better, if they pursue the exploitation of scale economies, fostering foreign direct investment and

promoting management education (Wennekers et al., 2005). Thus, so far in the literature there seems

to be a lack of empirical research of the role played by technological drivers on entrepreneurial

activity, at an aggregate level.

Acs et al. (2006) suggest that the differentiating factor of an entrepreneurial economy is the way

how entrepreneurs are used to facilitate knowledge spillovers. This kind of spillover and knowledge

commercialisation can be inhibited through the ‘knowledge filter’ (Audretsch & Stephan, 1999;

Audretsch & Lehmann, 2005). Acs et al. (2006) as well as Audretsch et al. (2006) argue that

entrepreneurship promotes economic growth, by permeating the ‘knowledge filter’ and

commercialising ideas declined by established firms.

FDI is defined as investment made to acquire a lasting interest in a firm operating in an economy

other than that of the investor. FDI is intimately linked to the activities of multinational corporations

(MNCs). The attraction of FDI plays an important role in public policies for entrepreneurship

promotion since MNCs have been argued to impact positively on indigenous entrepreneurial activity

(Acs et al., 2006). FDI is also associated with technology transfer and knowledge spillovers,

channelled through product and process technology, management practices (Findlay, 1978; Dyker,

1999), information about access to foreign countries (Rasiah, 1995) and intensified competition

(Blomström & Kokko, 1997; Markusen & Venables, 1999). Several authors have argued that the

economic activity of a foreign investor will help to accelerate technological development in the host

economy to some degree (Hunya, 2000; Lim, 2001; Dyker & Stolberg, 2003; Barbosa & Eiriz,

2007).

During the second half of the last century, the services industries played an increasingly

important role in the introduction of technological innovations. As a consequence of this, several

economists focused their research on the framework of technological change (Barras, 1986;

Andersen et al., 2000; Metcalfe & Miles, 2000; Devezas, 2005). Studies in the neo-Schumpeterian

literature, namely, Gershuny & Miles (1983), Barras (1990), Evangelista (2000) and Miles (2005)

consider ICTs to be a technological driver which promotes economic growth through a dynamic

10

process of creative destruction based on innovation. Services – such as, for instance, banking,

insurance and transport – are considered to be the main adopters of new technologies – especially

ICTs – which, through enhanced productivity, are the main engine of their increasing contribution

for national economic growth (OECD, 2000).

There is a growing literature about technological innovation in services that is grouped in three

approaches (Gallouj, 2002): (i) technologist (in which is stated that the introduction and diffusion of

ICT may improve productivity and economic performance); (ii) service oriented (which advocates

that innovation in manufacturing and services industries are different); and (iii) integrative (that

explores the boundaries between goods and services, and try to fill up the gap between them). In this

paper we follow a technologist approach, arguing that ICTs promote economic growth and act as a

technological driver that ‘pulls’ both technological and non-technological innovations associated

with FDI. Although certainly other factors are also essential sources of spillovers that foster

entrepreneurial activity, in this paper we concentrate on the investment in ICTs and its pull effect on

inward FDI.

THE IMPACTS OF FDI AND ICT INVESTMENT

In this section we review the prior empirical research, establishing a bridge between two

frameworks: (i) the ‘traditional’ focus on FDI as a source of spillovers; and (ii) the more recent focus

on ICTs. The first framework is concerned with the relationships established between FDI, trade,

economic development and entrepreneurship. The second deals with the impact of the technological

driver ICT on economic growth, employment and productivity. From the review of the two

frameworks, three hypotheses are derived.

FDI Framework

Fontagné & Pajot (1997) conducted empirical research on France, Sweden, and the United States,

revealing complementarities between trade and inward FDI. In the case of France and the US,

outward FDI is found to be a complement for exports, but substitute for imports. Inaba (1999)

analysed the effect of FDI on the Japanese balance of payments. The results revealed that, on the one

hand, FDI did not necessarily contribute to reducing the huge Japanese trade surplus, and on the

other hand, the worldwide structural changes may have had a great impact on the trade balance.

Ericsson & Irandoust (2001) used a Vector Autoregression (VAR) model in order to detect the

relationship between FDI and economic growth in Denmark, Finland, Sweden and Norway. The

11

authors found the existence of a feedback relationship in Sweden and a unidirectional relationship in

Norway. For Denmark and Finland no causality relationships were detected. The results were

justified through the existence of a larger number of MNCs, in Finland, especially, in services

industries.

De Backer & Sleuwaegen (2003) studied the relationship between FDI and domestic

entrepreneurship, and their findings are in line with theoretical occupational choice models that

predict FDI would crowd out domestic entrepreneurs through their selection in product and labour

markets. Nevertheless, empirical work has also found that the referred crowding effect may be

moderated or even reversed in the long run due to the long term positive effects of FDI on domestic

entrepreneurship as a result of experience, learning, demonstration and networking effects between

foreign and domestic firms (Rodriguez-Clare, 1996; Markusen & Venables, 1999).

A significant work by Gani & Sharma (2003) has shown that diffusion of new ICT instruments

are major pull factors of FDI.

Chang (2005) analyses the dynamic relationships among FDI, economic growth, unemployment

and trade in Taiwan, by making use of a Co-integrated Vector Autoregressive (CVAR) approach.

The results pointed out that both economic growth and exports have positive impacts on inward FDI,

whereas the expansion of exports impacts negatively on outward FDI. Other significant empirical

finding pointed out that there is no relationship between inward FDI and unemployment.

Additionally, a positive relationship exists between economic growth and exports, and a negative one

exists between unemployment and economic growth.

Barbosa & Eiriz (2007), by using firm-level panel data for the Portuguese manufacturing and

services industries, for the period 1986 – 2000, analyse the conditions of whether FDI had a positive

impact on entrepreneurial activity or not, by using as measure the net creation of firms. The authors

reveal that the impact of first foreign investment is, in general, positive, but that the marginal impact

of additional investments appears to be negative. One of the limitations revealed by the authors

constitutes an important guideline for the present research: weak evidence on positive effects of

MNCs on entrepreneurial activity might hide a relevant action of MNCs, as levers of technological

development, on industrial reorganisation.

ICT Framework

During the 1990s ICTs contributed in a remarkable way to productivity growth in several

economies (Vu, 2004). For example, the US labour productivity revived with a significant

acceleration during the period 1995-2000, and ICT investment accounted for more than one fifth of

Gross Domestic Product (GDP) growth throughout the last decade of the 20th century (Jorgenson &

Stiroh, 2000; CEA, 2001; Oliner & Sichel, 2001). The impact of ICTs on growth was also significant

12

in Australia (Parham et al., 2001), Canada (Armstrong et al., 2002), Korea (Kim, 2002), United

Kingdom (Oulton, 2002) and the Netherlands (Van der Wiel, 2001).

Antonelli (1998) analysed the co-evolution of ICTs and the knowledge intensive industries. The

results revealed that ICT affect the actual conditions of information, in terms of their basic

characteristics of appropriation and tradability, by favouring the role of business services as forces of

interaction amongst knowledge components in the generation of new technologies. Gretton et al.

(2002), studying firm-level data from the Australian Business Longitudinal Survey, found positive

and significant links between the use of ICTs and growth in both manufacturing and services

industries.

Pilat & Wolf (2004) examined the role of ICT producer and key ICT consumer sectors in

explaining overall productivity growth in OECD countries; they found that the impact of ICT

producer sectors is most significant in Finland, Ireland, and Korea, whereas ICT consumer sectors in

some countries, such as, the US and Australia, had an impressive growth in the second half of the

90s. Hempell et al. (2004) analysed comparable panel data of the Dutch and German firms in the

services industries and found that broadening the intensity of ICT capital and innovation have a

complementary impact on productivity.

Leitão & Ferreira (2008) analyse the impact of the liberalisation of European

Telecommunications Markets on the Business Ownership Rate. In the case of Germany, the Gross

Domestic Product precedes decreasing Business Ownership Rates, whereas, in the case of Portugal,

the Business Ownership Rate pulls for additional investments in ICT. Besides, a creative

entrepreneurial destruction is somehow ratified, since the Business Ownership Rate impacts,

negatively, on the level of employment.

Hypotheses

In this context, the analysis of the relevance of the Schumpeterian mechanism of creative

destruction through entrepreneurial activity arising from spillovers originating from the combination

of inward FDI and ICT investment requires the examination of three different hypotheses regarding

the causality relationships established between these two variables, and between them and

entrepreneurial activity:

• Hypothesis 1 (H1): Inward FDI causes the level of entrepreneurial activity in the host

economy;

• Hypothesis 2 (H2): Investment in ICT causes the level of entrepreneurial activity in the

host economy;

13

• Hypothesis 3 (H3): Inward FDI and investment in ICT jointly cause entrepreneurial

activity in the host economy.

DATA AND METHODOLOGY

Research about the aggregate impact of inward FDI and ICT investment on the level of

entrepreneurial activity of the host economies is, as far as we can tell, inexistent. To address this

caveat in the literature a comparative analysis between a high-tech producer (Finland) and a low-tech

producer (Portugal) is performed. The selection of these two European countries is justified by two

reasons. First, during the period of analysis, we observed in both countries an evolution from a

market dominated by one telecommunications operator totally or partially owned by the government

towards a liberalized market. This is particularly interesting for observing the impact of the end of

institutional monopolies and, as a consequence, the entry of new competitors on investment. Second,

both countries were affected, at different points in time by a major exogenous shock that changed

political, economic and regulatory governance. Portugal entered the European Community in 1986,

whereas in the case of Finland became a member in 1995.

Three databases are used covering the period 1976–2002. From COMPENDIA 20024, information

about Business Ownership Rate (BOR), Gross Domestic Product (GDP) and Unemployment

(UNEMP) was collected. The BOR is given by the ratio between number of business owners and

total employment. UNEMP is given by the ratio between the number of unemployed persons and

total workforce.

The data on inward Foreign Direct Investment (FDI) were collected from UNCTAD 20055. This

variable is defined by UNCTAD as an investment involving a long term relationship and reflecting a

lasting interest in, and control by, a resident entity in one economy (foreign direct investor or parent

enterprise) of an enterprise resident in a different economy (FDI enterprise or affiliate enterprise or

foreign affiliate). Such investment involves both the initial transaction between the two entities and

all subsequent transactions between them and among foreign affiliates. For statistical purposes, FDI

is typically defined as an incorporated or unincorporated enterprise in which the direct investor,

resident in another economy, owns 10% or more of the ordinary shares of voting power (or the

equivalent). However, this criterion is not strictly observed by all countries reporting6.

4 The data set is called COMPENDIA, which means COMParative ENtrepreneurship Data for International Analysis. It presents a harmonized data set over the period 1972-2002 for 23 OECD countries. It has been constructed making use of OECD statistics as well as other relevant sources (Van Stel, 2005). 5 The United Nations Conference on Trade and Development (UNCTAD) compiles statistics on FDI. The data are presented in two separate products: an interactive database for the aggregate figures and an electronic publication with the detailed information by country (http://stats.unctad.org/fdi/). 6 For detailed information please consult: www.unctad.org.

14

The data on investment in Information and Communication Technologies (IICT) were collected

from the ITU World Telecommunications Indicators 20067. According to the definition considered in

this database, this variable includes the total amount of investment in ICTs made both by public and

private agents in an economy.

The CVAR approach employed in the present study provides the possibility of carrying out

longitudinal studies and identifying long term economic relationships (Juselius, 2007). After

reviewing the empirical evidence, the econometric methodology follows an outline of four sequential

steps: (i) selection of an initial model specification; (ii) study of the integration order of the variables;

(iii) estimation process of the CVAR model; and (iv) dynamic analysis.

The Initial Model Specification

The BOR is used for measuring entrepreneurship in the studies by Audretsch and Thurik (2001),

Carree & Thurik (2006), Van Stel (2006), and Leitão & Ferreira (2008). Unemployment and Gross

Domestic Product are also considered in the initial model specification.

The VAR model applied to the cases of Finland and Portugal presents as differentiating element

the simultaneous inclusion of the variables related to the two determinants of entrepreneurial activity

being studied here: inward FDI and investment in ICTs.

The initial model specification is represented through a system of five equations by considering

five endogenous variables:

⎥⎥⎥⎥⎥⎥

⎦

⎤

⎢⎢⎢⎢⎢⎢

⎣

⎡

+

⎥⎥⎥⎥⎥⎥⎥

⎦

⎤

⎢⎢⎢⎢⎢⎢⎢

⎣

⎡

×

⎥⎥⎥⎥⎥⎥⎥

⎦

⎤

⎢⎢⎢⎢⎢⎢⎢

⎣

⎡

Ω

Ω

Ω

Ω

Ω

+

⎥⎥⎥⎥⎥⎥

⎦

⎤

⎢⎢⎢⎢⎢⎢

⎣

⎡

=

⎥⎥⎥⎥⎥⎥

⎦

⎤

⎢⎢⎢⎢⎢⎢

⎣

⎡

−

−

−

−

−

t

t

t

t

t

pt

pt

pt

pt

pt

t

t

t

t

t

t

t

t

t

t

uuuuu

IICT

FDI

UNEMP

GDP

BOR

IICTFDIUNEMPGDPBOR

5

4

3

2

1

4,53,5

4,43,4

4,33,3

4,23,2

4,13,1

2,5

2,4

2,3

2,2

2,1

1,5

1,4

1,3

1,2

1,1

5

4

3

2

1

θθθθθ

σσσσσ

βββββ

ααααα

(1)

where: tBOR , tGDP , tUNEMP , tFDI and tIICT are the variables representing Business Ownership

Rate; Gross Domestic Product; Unemployment; Foreign Direct Investment; and Investment in ICTs.

The number of lags is given by: kp ,...,1= , where k corresponds to the optimal number of lags ( )maxp ; t

corresponds to the year; and itu are the errors or the random disturbances.

7 The International Telecommunications Union (ITU) publishes The World Telecommunication/ICT Indicators Database that contains data from 1975 to 2006 for around 100 communications statistics covering telephone network size and dimension, other services, quality of service, traffic, staff, tariffs, revenue and investment (http://www.itu.int/publ/D-IND-WTID-2007/en).

15

The Integration Order of the Variables

The first step in the determination of the kind of relationship between the variables under study is

the application of unit root tests. First, we will evaluate if the time series are integrated or not, and,

then, if so, we will determine the integration order of the variables, in order to find the best way of

making it stationary8.

The procedures for detecting the existence of a unit root make use of the Dickey-Fuller (DF) Test

and of the Augmented Dickey-Fuller Augmented (ADF) Test (Dickey and Fuller, 1979). In order to

specify the model which provides the best adjustment, we make use of the Akaike Information

Criteria (AIC) and of the Schwarz Bayesian Criteria (SBC). For detecting error autocorrelation, the

LM test is used, and the probability of the Q statistics, originally, proposed by Ljung and Box (1979),

is also computed, taking into consideration the correlogram generated from the estimation process.

Differentiation of the time series showed that all the variables are stationary and integrated of order

one (Table 1 and Table 2).

Table 1 The ADF tests and the PP tests, including a constant, and without tendency

Variables

First Differences Finland Portugal ADF ADF

BOR -3,723*(5) -5,341*

GDP -4,801* -4,473*

UNEMP -4,819* -3,878*

FDI -3,202* -4,458*

IICT -4,645* -7,738*

* It denotes the rejection of the null hypothesis that is related to the existence of a unit root. + The number of lags is presented under brackets, when it is not automatically specified.

Table 2 The ADF tests and the PP tests, including a constant and tendency

Variables

First Differences Finland Portugal ADF ADF

BOR -4,009* -5,206*

GDP -4,718* -4,154*

UNEMP -4,720* -3,772*

FDI -3,923* -4,663*

IICT -4,784* 1,603

* It denotes the rejection of the null hypothesis that is related to the existence of a unit root.

The Estimation Process of the CVAR Model

In the selection process of the optimal number of lags (pmax), the values of five different

information criteria were computed. After detecting the inexistence of error autocorrelation, through

8 For information about selected descriptive statistics and correlation coefficients of the variables, see Appendix.

16

the use of Lagrange Multiplier (LM) tests, and considering the results obtained through all the

criteria, we retain that in the estimation process two lags should be considered (Table 3)9.

The analysis of error autocorrelation was carried out using the simulation of two different

estimation processes, and by making use of LM tests. For both cases, two lags were considered in the

estimation of VAR models. In order to detect the number of cointegration relationships, we follow

Johansen and Juselius (1990). The principle of the maximum likelihood is taken into consideration,

by using the Trace Statistic and the Max-Eigenvalue Statistic (Table 4).

Table 3 The selection process of the optimal number of lags

* It identifies the optimal number of lags selected through each one of the information criteria.

Table 4 The Cointegration Tests

Hypotheses Traceλ Hypotheses Maxλ

EV H0 H1 Observed Critical H0 H1 Observed Critical

Finland

0.955797 r=0 r=1 122.3235* 69.81889 r=0 r>0 74.85523* 33.87687 0.671868 r=1 r=2 47.46830 47.85613 r≤ 1 r>1 26.74414 27.58434

Portugal 0.971734 r=0 r=1 162.4114* 69.81889 r=0 r>0 85.58602* 33.87687 0.870264 r=1 r=2 76.82540* 47.85613 r≤ 1 r>1 49.01417* 27.58434 0.518193 r=2 r=3 27.81123 29.79707 r≤ 2 r>2 17.52506 21.13162

[+] The first column corresponds to the Eigenvalues (EV).; [++] The critical values of the Trace Statistic and of the Max-Eigenvalue Statistic, at a 5% significance level, were collected from Osterwald-Lenum (1992); * It denotes the rejection of the initial hypothesis, at a 5% significance level.

According to the results previously displayed in Table 4, we reject the null hypothesis of

nonexistence of cointegration relationships among the variables. From the remaining lines of the test,

we conclude that, in the case of Finland, just one co-integrating vector is required, whereas in the

case of Portugal, two co-integrating vectors are considered in the subsequent estimation process of

the CVAR model. Following Engle & Granger (1987) the use of Error Correction Terms (ECT) aims

to assure that the disequilibrium proportion observed in the set of variables, in a certain period, is

corrected in the next one10.

9 For a discussion about the use of different information criteria, see Lütkepohl (1999, 2004). 10 According to Engle & Granger (1987), an example of this is given by the fact that a change in prices, in a certain period, can result from excess demand on the previous period.

Lags LR FPE AIC SBC HQ Finland 0 NA 4.22e+25 73.19478 73.43856 73.26239 1 140.3517 2.03e+23 67.80785 69.27050 68.21353 2 77.43235* 7.81e+21* 64.27697* 66.95850* 65.02071* Portugal 0 NA 1.10e+28 78.75701 79.00079 78.82463 1 108.8216 2.78e+26 75.02956 76.49221 75.43524 2 68.68361* 2.00e+25* 72.12359* 74.80511* 72.86733*

17

In the present analysis, for the Finnish case, the ECT1 term is incorporated into the specification

of the model to be tested, whereas in the case of Portugal, two more terms are considered for

accomplishing the same aim, namely, the ECT2 and the ECT3.

RESULTS AND DISCUSSION

Following Granger (1969), a dynamic analysis based on the evaluation of the causality

relationships was performed (Table 5). In addition, the results regarding feedback causality

relationships are discussed by making use of the Cholesky’s variance decomposition of the

forecasting error and of simulated coefficients of impulse-response functions (Table 6).

Table 5 The Contrasts of the Granger Causalities

Dependent BORΔ GDPΔ UNEMPΔ FDIΔ IICTΔ

Independent Finland

BORΔ - 0.460220 0.433286 10.39003* 1.099391 GDPΔ 3.805439 - 2.436646 18.48589* 6.684646*

UNEMPΔ 5.409202** 1.615022 - 20.36600* 0.052119 FDIΔ 6.175823* 4.287294 8.003672* - 49.56665* IICTΔ 8.396432* 6.307559* 3.832082 8.650869* -

Block 11.53676 15.36291** 22.54837* 26.59668* 84.97624* ECT1 0.176836 -31102.23 105.4301 687226.6* -3.08E+10

Portugal

BORΔ - 0.516053 2.815914 11.99136* 3.133633 GDPΔ 1.503691 - 14.45955* 13.84465* 10.39995*

UNEMPΔ 2.195985 0.696415 - 4.442577 6.315380* FDIΔ 0.725456 2.023997 34.68989* - 18.77583* IICTΔ 2.911339 3.140926 42.02979* 28.75726* -

Block 7.674842 4.767031 68.36573* 48.91318* 21.01871* ECT2 -0.904856 13906.42 -48.98935 -177494.8 1.21E+13 ECT3 4.20E-05 0.228865 -0.006116* 13.82023* 5.74E+08*

[+] The contrasts of the causality of the variables are made by using the2χ statistic, with one degree of freedom, while the contrasts of the significance of the Error

Correction Terms (ECT) are made through the use of the t statistic. * Significance level: 5%. ** Significance level: 10%.

Finland

In the case of Finland, the results provide the identification of feedback causality relationships

comprising the following pairs of variables: (FDI, BOR), (IICT, GDP), (FDI, UNEMP) and (IICT,

FDI).

In what concerns Hypothesis 1 (H1), the pair (FDI, BOR), after two years, FDI does not present a

significant impact on the BOR (its weight is less than 5%). Nevertheless, after the fifth year, FDI

starts to have persistently growing significance on the determination of the BOR. The sum of the first

ten coefficients obtained through the use of the forecasting technique of impulse-response functions,

revealed a negative signal for the relationship between the FDI and the BOR. Thus, by considering

18

the summing up of the coefficients, we may state that the bigger the FDI is, the smaller the BOR will

be in case of Finland. On the contrary, the result obtained for the causality relationship established

from the BOR towards FDI is extremely important, since it suggests that promoting an

entrepreneurial and dynamic environment contributes to attract further FDI. According to the results

obtained, the BOR impacts positively and significantly on inward FDI, by explaining 17.65% of its

forecasting error after two years.

Table 6 Dynamic Analysis of the Feedback Causalities: Finland vs Portugal

Pairs of Variables Feedback Causalities Dynamic

Analysis 2 Years 3 Years 10 Years

Signal of the Relationship

Finland

(FDI, BOR) BORFDI Δ→Δ *

VDC 2,42E-05 1,815 7,483 - IRF -1,56E-06 -0,0005 -0,005

FDIBOR Δ→Δ * VDC 17,654 13,188 14,164

+ IRF 1002,636 520,526 333,424

(IICT, GDP) GDPIICT Δ→Δ *

VDC 0,129 3,321 5,136 + IRF -14,345 74,739 555,212

IICTGDP Δ→Δ * VDC 7,696 15,456 36,087

- IRF -1,60E+08 -5,10E+08 -4,07E+09

(FDI, UNEMP) UNEMPFDI Δ→Δ *

VDC 2,081 9,326 7,241 + IRF 0,192 0,667 2,260

FDIUNEMP Δ→Δ VDC 0,826 0,719 2,891

- IRF -267,294 -127,125 -775,012

(IICT, FDI) FDIIICT Δ→Δ

VDC 1,858 1,710 1,461 + IRF -307,669 -79,338 433,639

IICTFDI Δ→Δ * VDC 84,22 69,30 47,58

- IRF -8,85E+08 -1,41E+09 -5,00E+09 Portugal

(IICT, UNEMP) UNEMPIICT Δ→Δ

VDC 0,030 0,956 1,288 - IRF 0,010 -0,093 -1,132

IICTUNEMP Δ→Δ * VDC 13,720 9,451 16,255

+ IRF 3,59E+10 4,62E+10 6,15E+10

(IICT, FDI) FDIIICT Δ→Δ

VDC 0,051 0,113 1,515 + IRF 18,511 46,024 797,033

IICTFDI Δ→Δ * VDC 3,042 1,966 13,087

+ IRF 1,95E+10 1,65E+10 1,15E+10 Legend: VDC is the Variance Decomposition of Cholesky; IRF corresponds to the Impulse-Response Functions. * It is significant when the impact is higher than 5% (Goux, 1996). [+] The sign of the percentage weight is obtained through the sum of the coefficients of the first 10 periods (Goux, 1996).

The pair (IICT, GDP) also provides interesting results in the sense that, on one hand, IICT has a

positive impact on GDP, although it should be considered a long term impact because it only

becomes significant from the fifth year onwards; on the other hand, the GDP impacts in a negative

way on the behaviour of the IICT variable. It presents a significant and growing impact starting from

the second period. This is justifiable by decreasing levels of IICT in more developed countries that

were observed during the 1980s, and also in the first half of the 1990s.

With respect to the pair (FDI, UNEMP), it should be stressed that FDI impacts positively on the

UNEMP variable. This is an expected result, especially if we are dealing with foreign investment that

leads to the downsizing of existing local firms and promotes efficiency gains through competitive

19

pressure. The other side of the feedback causality relationship reveals that UNEMP does not have a

significant impact on the FDI.

In what concerns Hypothesis 2 (H2), and taking into consideration the results obtained in terms of

unidirectional causalities, it should be stressed that IICT impacts the behaviour of BOR negatively,

but not significantly. Additionally, the GDP has a positive and significant impact on FDI. This result

is quite important, since it reveals how the creation of wealth is an effective mechanism of signalling

that makes possible the attraction of FDI.

In what concerns Hypothesis 3 (H3), the pair (IICT, FDI), IICT has a positive impact on inward

FDI, which is not significant in the case of Finland. Inward FDI negatively the IICT, and it

contributes for explaining 84.22% of the variance decomposition of IICT’s forecasting error, which

provides further insights for the aforementioned decreasing levels of IICT during the 1980s and the

first half of 1990s. This result also provides useful insights about the way the combination of

investment in ICTs and inward FDI impacts on the dynamics of the entrepreneurial activity in the

host economy.

Portugal

In the case of Portugal, the results from the dynamic analysis do not provide any significant

support for both Hypothesis 1 (H1) and Hypothesis 2 (H2). In fact, it is only possible to detect that the

GDP (starting from the first period) and the FDI (starting from the eighth period) have a significant

and positive impact on the behaviour of the UNEMP variable.

Nevertheless, the BOR and the GDP have a significant impact on the FDI variable. On the one

hand, the BOR has a positive impact on the FDI, which reveals the importance of promoting a

dynamic and entrepreneurial environment, in order to attract additional FDI; on the other hand, the

GDP has a negative impact on the FDI, although this causality relationship has a different sign from

the second period until the eighth period of the simulation of impulse-response coefficients. Finally,

the GDP has a negative impact on the IICT, as it was previously detected in the case of Finland,

although the impulse of one standard deviation innovations reveal the existence of a much reduced

impact on the behaviour of the IICT variable.

In terms of feedback causality relationships, two pairs of variables were detected, namely: (IICT,

FDI) and (IICT, UNEMP). In what concerns Hypothesis 3 (H1), the first pair of variables provides

interesting insights, since IICT impacts positively, although not very significantly, on FDI. This is a

very important result, since the argument goes that, for attracting additional values of inward FDI,

increasing values of IICT should be ensured. This is a quite important feedback relationship, since

FDI has a positive effect on IICT, explaining 9.55% of the forecasting error of IICT after just one

20

year. In the fourth period it presents a positive impact of 11.05% of the variance decomposition of

the forecasting error of IICT.

Although the causality relationship established from IICT to UNEMP presents a negative signal,

this does not represent a significant impact, at least for a forecasting period of ten years. There is,

however, a positive and significant impact of UNEMP on IICT, which is probably due to increasing

public investment in ICTs, during the first half of 90s, a period of low economic growth and

increasing unemployment. Government policies dealing with the long term causes of unemployment

included the improvement of infra-structure and education, strongly associated with investment in

ICTs.

CONCLUDING REMARKS

This study has provided evidence that, in disagreement with expectations, both inward FDI and

ICT impact negatively on business ownership rates. This result suggests that, while FDI and

investment in ICTs may represent sources of spillovers in a Schumpeterian process of creative

destruction, it is likely that these spillovers benefit mostly existing and not new firms. Moreover, any

positive effects that these two variables may have on entrepreneurial activity are likely surpassed by

an effect of displacement of existing firms by MNCs or by competitive new firms, leading to a net

effect on the business ownership rate that is negative. In the case of FDI, under a neo-Schumpeterian

approach, the exit of entrepreneurs in the host economy reveals the pressure for eliminating the more

inefficient (or less innovative) entrepreneurial units that is provided by the entry of MNCs, which

follow innovative conducts that tend to reinforce the entrepreneurial innovative capacity of the host

economies.

A possible reason for the result is the metric used for entrepreneurial activity: the business

ownership rate, measured as the proportion of business owners in the labour force. This measure

does not gauge entrepreneurial activity as consistently as, for instance, the number of start-ups in the

economy. Also, in particular for the case of Portugal, business ownership rates tend to include a high

proportion of necessity-based, subsistence entrepreneurial activity, which is not really innovative and

oriented towards growth, but simply an occupational choice derived from the lack of better

alternatives (Baptista and Thurik, 2007). If a large proportion of entrepreneurial activity is of this

kind, than business ownership rates are likely to respond positively to increases in unemployment

and reductions in the amount of job opportunities in the economy. Inversely, the increased job

opportunities that are typical of a growing economy experiencing high levels of FDI (and ICT

investment) will likely yield reductions in business ownership rates, as individuals leave subsistence

entrepreneurship because they found better alternatives in wage employment.

21

In the Finnish case, FDI plays a particular important role as a determinant of domestic

entrepreneurship, since it contributes significantly for decreasing values of entrepreneurial activity.

This may be due to large average firm size and the significant entry barriers that are associated with

the dominant technological profile of the high-tech activities specialisation in this economy. Also,

the creation of high-productivity, high wage jobs by increased levels of FDI in these sectors

increases the opportunity cost of entrepreneurship.

The feedback causality between investment in ICTs and FDI revealed a positive impact of FDI

on investment in ICTs, in the case of a low tech producer (Portugal), and a negative impact of FDI on

investment in ICTs in the case of the high tech producer (Finland). This feedback causality

relationship provides important insights for public policy makers about the role played by the inward

FDI, as a technological driver, in terms of the determination of future investments in ICTs, in

countries with different experience curves and technological profiles.

Revealing this contrasting result is very important since, in the case of Finland, the

implementation of public policies oriented to inward FDI precedes decreasing values of investment

in ICTs, which reveals the growing importance of the private sector in ensuring investment in this

technological driver, which was previously supported by public funding. In the case of Portugal, FDI

plays an important role, in the long term, since it impacts positively on investment in ICTs which are

the catalyst driver for promoting a sustainable process of technological change. This justifies the

design of further public policies oriented to the attraction of FDI in order to modify the technological

profile of a small dimension nation and to revitalize its entrepreneurial innovative capability, through

a technological process of creative destruction based on the creation of further SMEs networked with

MNCs.

Another interesting result provided by the comparative analysis performed is with regard to the

impact of FDI on unemployment. In the case of Finland, there is a negative impact which is

significant starting from the third period, and the positive signal never changes during the simulation

for the next ten periods, suggesting that, after an initial effect of increased unemployment, possibly

due to increased competition and displacement of local firms, FDI serves as sustained a boost for

competitiveness and economic growth. The case of Portugal is quite different. While in the first

period the impact is null, and in the next two periods the effect is negative, afterwards the signal

becomes positive. Hence, for the Portuguese economy, inward FDI seems to contribute for

decreasing levels of UNEMP in the short run, possibly through direct effects associated to new jobs

created by MNCs, but this effect is not sustained over time, and seems to fluctuate depending on the

internal policies of MNCs regarding local investment and divestment.

The present study has four limitations. First, no measure of each country’s technological profile

was included in the analysis due to data constraints. Second, the use of aggregate data does not allow

for the examination of potential spillovers generated through networks of MNCs and high-tech

22

SMEs. Third, only two European countries are included in the comparative analysis. Further research

is needed in a European context, for contrasting the role played the dynamic drivers of

entrepreneurial activity. Fourth, the role played by human capital and organizational capital is not

explored in the present analysis, although it seems important to study how they act as determinants of

the ‘allocation’ of entrepreneurship, under an internationalisation context.

ACKNOWLEDGEMENTS

Valuable comments on an earlier draft of this paper were provided by several participants in the RENT XXI

Conference, Cardiff, in November 2007. A special debt of gratitude is due to Markku Virtanen, Jarkko

Pellikka, Tommaso Minola and Luca Iandoli for their support and positive incentives in this entrepreneurial

venture. We are also grateful to two anonymous reviewers of Inter-RENT-2008, for their comments.

REFERENCES

Acs, Z. (2006) How is Entrepreneurship Good for economic Growth?. Innovations, 1(1), 97-107. Acs, Z. and Storey, D. (2004) Introduction: Entrepreneurship and Economic Development. Regional Studies,

38(8), 871-877. Acs, Z. and Varga, A. (2005) Entrepreneurship, Agglomeration and Technological Change. Small Business

Economics, 24(3) 323-334. Acs, Z. and Szerb, L. (2006) Entrepreneurship, Economic Growth and Public Policy. Small Business

Economics, 28, 109-122. Acs, Z. and Armington, C. (2006) Entrepreneurship, Geography and American Economic Growth.

Cambridge: Cambridge University Press. Acs, Z., Audretsch, D., Braunerhjelm, P. and Carlsson, B. (2006) The Knowledge Spillover Theory of

Entrepreneurship. Centre for Economic Policy Research, Discussion Paper. Andersen, B.; Howells, J.; Hull, R.; Miles, I. and Roberts, J. (2000) Knowledge and Innovation in the New

Service Economy. Edward Elgar, Cheltenham. Antonelli, C. (1998) Localized technological change, new information technology and the knowledge-based

economy: The European evidence. Journal of Evolutionary Economics, 8, 177-198. Armstrong, P., Harchaoui, T., Jackson, C. and Tarkhani, F. (2002) A Comparison of Canada-US Economic

Growth in the Information Age, 1981 - 2000: The Important of Investment in Information and Communication Technologies. Economic Research Paper Series No. 70, February, Zurich.

Audretsch, D. and Stephan, P. (1999) How and why does knowledge spills over in biotechnology?. In: Audretsch, D., Thurik, R. (Eds.), Innovation, Industry Evolution, and Employment. Cambridge University press, Cambridge, 216–229.

Audretsch, A. and Thurik, R. (2001) Linking Entrepreneurship to Growth. OECD Science, Technology and Industry Working papers, 2001/2, OECD Publishing.

Audretsch, D. and Lehmann, E. (2005) Does the Knowledge Spillover Theory of entrepreneurship hold for regions?. Research Policy, 34, 1191-1202.

Audretsch, D.; Keilbach, M. and Lehmann, E. (2006) Entrepreneurship and Economic Growth. New York: Oxford University Press.

Baptista, R. and Thurik, R. (2007) The relationship between entrepreneurship and unemployment: Is Portugal an outlier?. Technological Forecasting & Social Change, 74, 75–89.

Baptista, R., Escária, V. And Madruga, P. (2008) Entrepreneurship, regional development, and job creation: the case of Portugal. Small Business Economics, 30, 49-58.

Barbosa, N. and Eiriz, V. (2007) The role of inward foreign direct investment on entrepreneurship. International Entrepreneurship and Management Journal (published online, May 2).

Barras, R. (1986) Towards a theory of innovation in services. Research Policy, Vol.15, 161 – 73. Barras, R. (1990) Interactive innovation in financial and business services: the vanguard of the services

Revolution. Research Policy, Vol.19, 215 – 237.

23

Baumol, W. (1968) Entrepreneurship in Economic Theory. The American Economic Review, Vol. 58, N.º 2, May, 64-71.

Baumol, W. (1990) Entrepreneurship: Productive, Unproductive, and Destructive. The American Economic Review, Vol. 98, N.º 5, October, 893-921.

Blomström, M. and Kokko, A. (1997) How Foreign Investment Affects Host Countries. Policy Research Working Paper, 1745, World Bank.

Burke, A., Gorg, H. and Hanley, A. (2007) The Impact of Foreign Direct Investment on New Firm Survival in the UK: Evidence for Static vs. Dynamic Industries. Small Business Economics (published online, July 5).

Carree, M. and Thurik, R. (2006) The lag structure of the impact of business ownership one economic performance in OECD countries. Discussion papers on Entrepreneurship, Growth and Public Policy, MPI Jena.

CEA (2001) Annual Report of the Council of Economic Advisers. The Economic Report of the President, January 2001, Council of Economic Advisors (CEA).

Chang, S-C. (2005) The dynamic interactions among foreign direct investment, economic growth, exports and unemployment: evidence from Taiwan. Economic Change and Restructuring, 38, 235-256.

De Backer, K. and Sleuwaegen, L. (2003) Does Foreign Direct Investment Crowd Out Domestic Entrepreneurship. Review of Industrial Organization, 22, 67-84.

Devezas, T. (2005) Evolution theory of technological change: State-of-the-art and new approaches. Technological Forecasting & Social Change, 72, 1137-1152.

Dickey, D. and Fuller, W. (1979) Distribution of the estimators for autoregressive Time series with Unit Root. Journal of the American Statistical Association, 74: 427–431.

Dyker, D. (ed) (1999) Foreign Direct Investment and Technology Transfer in the former Soviet Union. . Edward Elgar, Cheltenham.

Dyker, D. and Stolberg, C. (2003) Productivity and Capability in the Transition Economies: a Historical and Comparative Perspective. Brighton: University of Sussex, mimeo.

Engle, R. and Granger, C. (1987) Co-Integration and Error Correction: Representation, Estimation and Testing. Econometrica, 55, 251 – 276.

Ericsson, J. and Irandoust, M. (2001) On the Causality between Foreign Direct Investment and output: a Comparative Study. The International Trade Journal, Volume XV, No. 1, Spring.

Evangelista, R. (2000) Sectoral patterns of technological change in services, economics of innovation. Economics of Innovation and New Technology, Vol.9, 183–221.

Findlay, R. (1978) Relative Backwardness, Direct Foreign Investment, and the Transfer of Technology: A Simple Dynamic Model. Quarterly Journal of Economics, 92, 1-16.

Fontagné, L. and Pajot, M. (1997) How Foreign Direct Investments Affects International Trade and Competitiveness: an Empirical Assessment. Centre d’Études Prospectives et d’Informations Internetionales, Paris.

Fritsch, M. and Mueller, P. (2004) The Effects of New Business Formation on Regional Development over Time. Regional Studies, 38, 961-975.

Fritsch, M. and Mueller, P. (2008) The Effect of New Business Formation on Regional Development over Time: The Case of Germany. Small Business Economics, 30, 15-29.

Gallouj, F. (2002) Innovation in the Service Economy: The New Wealth of Nations. Edward Elga, Cheltenham. Gani, A. and Sharma, B. (2003) The Effects of Information Technology Achievement and Diffusion on

Foreign Direct Investment. Perspectives on Global Development and Technology, Volume 2, Issue 2, 161-178.

Gershuny, J. and Miles, I. (1983) The New Service Economy: the Transformation of Employment in Industrial Societies. London: Pinter.

Goux, J-F. (1996) Le canal étroit du crédit en France. Revue D’Économie Politique. 106(4), Juillet-Août, 1996, 655–681.

Granger, C. (1969) Investigating Causal Relations by Econometric Models and Cross Spectral Methods. Econometrica, 37, 424–438.

Gretton, P.; Gali, J. and Parham, D. (2002) Uptake and Impacts of ICT in the Australian Economy: Evidence from Aggregate, Sectoral and Firm Levels, OECD Workshop on ICT and Business Performance, Productivity Commission, Canberra, December.

Hempell, T.; Van Leeuwen, G. and Van der Wiel, H. (2004) ICT, Innovation and Business Performance in Services: Evidence for Germany and the Netherlands, In The Economic Impact of ICT—Measurement, Evidence, and Implications, Paris: OECD, 131 - 152.

Hunya, G. (ed.), (2000) Integration through Foreign Direct Investment. . Edward Elgar, Cheltenham.

24

Inaba, K. (1999) Japanese Foreign Investment: An Empirical study Using Multi-Sectoral Econometric Model. International Institute for Applied System Analysis, Interim Report – 031, Austria.

Johansen, S. and Juselius, K. (1990) Maximum Likelihood Estimation and Inference on Cointegration with Applications to Money Demand. Oxford Bulletin of Economics and Statistics, 52, 169–210.

Jorgenson, D. and Stiroh, K. (2000) Raising the Speed Limit: US Economic Growth in the Information Age. OECD Working Papers, No. 261.

Jovanovic, B. (1994) Firm Formation with Heterogeneous Management and Labor Skills. Small Business Economics, 6, 185-191.

Juselius, K. (2007) The Co-integrated VAR Model – Methodology and Applications. Oxford University Press. Kim, S. (2002) The Digital Economy and the Role of Government: Information Technology and Economic

Performance in Korea. Program on Information Resources Policy, Harvard University, January.

Leitão, J. and Ferreira, J. (2008) Entrepreneurship and ICT: A Comparative analysis between Germany and Portugal. International Journal of Entrepreneurship and Small Business, Forthcoming.

Lim, E.-G. (2001) Determinants of, and the Relation between, Foreign Direct Investment and Growth: A Summary of the Recent Literature. IMF Working Paper, WP/01/175, Washington DC: International Monetary Fund.

Ljung, G. and Box, G. (1979) On a Measure of Lack of Fit in Time Series Models. Biometrika, 66, 265 – 270. Lucas, R. (1993) Making a Miracle. Econometrica, 61(2), 251-272. Lütkepohl, H. (1999) Vector Autoregressions. Working Paper of Institut für Statistik und Ökonometrie,

Humboldt, Universitat, Berlin. Lütkepohl, H. (2004) Recent Advances in Cointegration Analysis. Economics Working Papers ECO2004/12,

European University Institute, Florence. Markusen, J. and Venables, A. (1999) Foreign Direct Investment as a Catalyst for Industrial Development.

European Economic Review, Vol. 43, 335-356. Metcalfe, J. and Miles, I. (2000) Innovation System in the Service Economy. Measurement and Case Study

Analysis. Boston: Kluwer. Miles, I. (2005) Innovation in services. In Fagerberg, J., Mowery, D.C. and Nelson, R.R. (eds.), The Oxford

Handbook of Innovation, Oxford: Oxford University Press. OECD (2000) Science, Technology and Innovation in the New Economy. Paris: OECD. Oliner, S. and Sichel, D. (2001) The Resurgence of Growth in the Late 1990s: is Information Technology the

Story?. Journal of Economic Perspectives, 14(4), 3-22. Osterwald-Lenum, M. (1992) A note with quantiles of the asymptotic distribution of the maximum likelihood

cointegration rank test statistics. Oxford Bulletin of Economics and Statistics, 54, 461–471. Oulton, N. (2002) ICT and productivity growth in the UK. Oxford Review of Economic Policy, Vol. 18, 363 -

379. Parham, D.; Roberts, P. and Sun, H. (2001) Information Technology and Australia’s Productivity Surge. Staff

Research paper, Productivity Commission, AusInfo, Canberra. Parker, S. (2005) The economics of Entrepreneurship: What we know and what we don’t. Foundations and

Trends in Entrepreneurship, 1(1), 1-54. Parker, S. (2006) The Economics of Entrepreneurship: What we know and what we don’t. Summer Institute of

the Entrepreneurship, Growth & Public Policy Group, Max Planck Institute of Economics, July of 2006, Jena, Germany, mimeo.

Pilat, D. and Wolfl, A. (2004) ICT Production and ICT Use: What role in Aggregate Productivity Growth?. In The Economic Impact of ICT-Measurement, Evidence, and Implications, 85 - 104, OECD, Paris.

Rasiah, R. (1995) Foreign Capital and Industrialisation in Malaysia. St Martin’s Press, New York. Rodriguez-Clare, A. (1996) Multinationals, Linkages, and Economic Development. American Economic

Review, Vol. 86 (4), 852-873. Schumpeter, J. (1940) Capitalism, Socialism and Democracy. Harper & Row, New York. Sleuwaegen, L. and Dehandschutter, W. (1991) Entry and Exit in Belgian Manufacturing. In P. Geroski and J.

Schwalbach (eds.), Entry and Market Contestability: An International Comparison, 111-121. Basil Blackwell, Oxford.

Van der Wiel, H. (2001) Does ICT Boost Dutch Productivity Growth?. CPB Document No. 016, CPB Netherlands Bureau of Economic Policy Analysis, December.

Van Praag, C., Cramer, J. (2001) The Roots of Entrepreneurship and Labor Demand: Individual Ability and Low Risk Aversion. Economica, 68(269), 45-62.

Van Stel, A. (2005) COMPENDIA: Harmonizing Business Ownership Data Across Countries and Over Time. International Entrepreneurship and Management Journal 1 (1), 105-123.

25

Van Stel, A., Carree, M. and Thurik, R. (2005) The Effect of Entrepreneurial Activity on National Economic Growth. Small Business Economics 24(3), 193-203.

Van Stel, A. (2006) Empirical Analysis of Entrepreneurship and Economic Growth. International Studies in Entrepreneurship, Springer.

Vu, K. (2004) Measuring the Impact of ICT on Economic Growth: An International Comparative Perspective. Workshop on “ICT as Drivers of Economic Development in Post-communist Countries”, Conference Board, New York, 19-20 January.

Wennekers, S., Van Stel, A., Thurik, R. and Reynolds, P. (2005) Nascent entrepreneurship and the level of economic development. Discussion Papers on Entrepreneurship, Growth and Public Policy, edited MPI Jena.

26

APPENDIX

Table A Selected Descriptive Statistics of the Variables for Finland

BOR GDP UNEMP FDI IICT Mean 0.072089 321.6823 0.233333 1746.582 1.99E+09 Median 0.075623 341.7458 -0.300000 406.7509 1.82E+09 Maximum 0.081754 871.8411 4.800000 12143.49 4.30E+09 Minimum 0.058854 -920.4193 -1.918378 -246.5381 7.74E+08 Std. Dev. 0.008421 441.2262 1.790152 3117.576 1.10E+09 Skewness -0.338145 -1.437614 1.240858 2.168159 0.619625 Kurtosis 1.466140 5.417612 3.963561 6.704628 2.174249

Jarque-Bera 3.161357 15.87575 7.973282 36.59391 2.494808 Probability 0.205835 0.000357 0.018562 0.000000 0.287250

Sum 1.946416 8685.422 6.300000 47157.72 5.38E+10 Sum Sq. Dev. 0.001844 5061695. 83.32076 2.53E+08 3.14E+19

Observations 27 27 27 27 27

Table B Correlation Matrix of the Variables for Finland

Variables BOR GDP UNEMP FDI IICT BOR 1 GDP 0.142645 1

UNEMP -0.134770 -0.904493 1 FDI 0.553026 0.366402 -0.309737 1 IICT 0.567294 0.132589 -0.128518 0.069988 1

Table C Selected Descriptive Statistics of the Variables for Portugal

BOR GDP UNEMP FDI IICT Mean -0.000470 198.0634 6.475926 1396.465 6.57E+10 Median -0.000643 191.1397 7.100000 921.5800 2.97E+10 Maximum 0.010469 471.1203 8.600000 6635.315 2.71E+11 Minimum -0.008915 -109.1587 4.100000 57.91000 1.58E+09 Std. Dev. 0.005066 175.5194 1.503737 1721.217 7.76E+10 Skewness 0.362685 -0.018478 -0.186934 1.896270 1.209056 Kurtosis 2.791856 2.101646 1.585934 6.234974 3.460427

Jarque-Bera 0.640671 0.909457 2.406778 27.95446 6.816669 Probability 0.725905 0.634620 0.300175 0.000001 0.033096

Sum -0.012678 5347.711 174.8500 37704.54 1.77E+12 Sum Sq. Dev. 0.000667 800983.7 58.79185 77027268 1.56E+23

Observations 27 27 27 27 27

27

Table D Correlation Matrix of the Variables for Portugal

Variables BOR GDP UNEMP FDI IICT BOR 1 GDP 0.034046 1

UNEMP -0.334594 -0.308468 1 FDI 0.124628 0.204099 -0.736967 1 IICT 0.424908 0.207737 -0.249285 0.182683 1

28

CREATING TECHNOLOGY-BASED COMPANIES – ABOUT THE SIGNIFICANCE OF EFFECTUAL LOGIC IN THE EARLY PHASE OF THE VENTURING PROCESS

René Mauer, Malte Brettel

RWTH Aachen University, Germany. Contact details: RWTH Aachen, LS WIN, Templergraben 64, 52053 Aachen, Germany. Tel.: ++49/241/8099-362. Fax.: -523. E-Mail: mauer@win.rwth-

aachen.de.

ABSTRACT

Studies in technology entrepreneurship acknowledge a high degree of uncertainty in connection with both the notion of technology and the early-stage process. However, only few studies have examined in detail how technology entrepreneurs encounter this uncertainty. Cognition-based literature on the other hand has come up with the theoretical concept of effectuation – a logic of non-predictive control that expert entrepreneurs apply successfully to overcome uncertainty. Our study presents empirical data on 8 case studies of early-stage technology ventures, differentiating causal and effectual approaches within the dynamic entrepreneurial process and connecting the findings with a performance measure that builds upon the concept of gestation steps. We find that variation among our set of technology entrepreneurs can be explained through varying sophistication levels of uncertainty management, which we define as the capability to combine effectual with causal behaviour with respect to the type of uncertainty that dominates the situation. The results suggest that a more contingent application of the toolboxes depending on the type of uncertainty in connection with teaching of effectual logic can add successfully to venture performance.

29

INTRODUCTION

Personal experience with technology entrepreneurship projects showed us an interesting variation

among the performance of technology ventures in their early phase, on their way to develop a

technology into a product and creating a company on this basis. Being confronted with this

phenomenon, we were not content with explanations that pointed to singular aspects like team

structures, the elaboration of business plans, etc. We asked two questions: Is the phenomenon of

technology entrepreneurship important enough to engage with research in the difficult to capture

early-stage process? And if so, what might explain the variation in terms of a more comprehensive

framework?

In 2007, Kirchhoff et al. provided us with latest results on university R&D expenditures and new

business formations, once again confirming a positive economic impact of technology-based

entrepreneurial activity. According to this study, university R&D expenditures are positively related

to the creation of new firms as well as to employment level and change (Kirchhoff et al., 2007).

Similar studies with different notions on how technology entrepreneurship reveals its positive impact

all support this basic claim that Schumpeter had referred to as the innovative power of the

entrepreneur in the middle of the last century (e.g. (Autio, 1997).

However valuable this economic activity seems to be, it also seems to be difficult to induce.

German policy makers, for example, worry about a decreasing number of technology-based

companies due to a strong job market for technology experts, displaying economic alternatives that

seem to carry lower levels of risk and uncertainty (Niefert et al., 2006). This current study once again

supports an old discussion – in this case Knight’s ideas about this trade-off by each individual which

he formulated in this work in 1921 (Knight, 1921).

There has been quite a lot of research on technology or technology-based entrepreneurship due to

the growing interest of policy makers interested in the economic benefits (Bell & Pavitt, 1993;

Bullinger, 2002; Hsu et al., 2007; Licht & Nerlinger, 1998; Shane, 2004a, 2004b). However, the

phenomenon is facing a paradox. On the one hand there is the desire to create new economic

artefacts (new firms) on the basis of newly developed technologies, implying a high level of

uncertainty on several dimensions. On the other hand, research on the creation process mainly draws

from the knowledge about established companies. Although, in general, it has been supported that

opportunity recognition and business planning procedures are key determinants of the early-stage

process (e.g. Ardichvili, Cardoso, Ray 2003; Hills 1995; Kraus, Schwarz 2007; Shane, Delmar

2004), both concepts fail in the face of high levels of uncertainty. They stop short of offering viable

explanations for the early stages of the venturing process where key issues are still unclear:

unidentified market need, a new technology which still has to be developed into a product, an

30

undefined market and an unknown customer group (Bygrave & Hofer, 1991; Diochon et al., 2007;

Klofsten, 2005; Mellewigt & Witt, 2002; Reynolds & Miller, 1992).

We continue to introduce the principles and the dynamic model of effectuation into that

discussion as behaviour based on non-predictive logic as opposed to prediction-based causal logic.

An effectual process mainly starts with a given means base and develops a project through a series of

commitments by self-selected stakeholders. That said, it offers ideas for the early phases of the

venturing process that do not require prediction of the future (S. D. Sarasvathy & Dew, 2005a). Our

study thereby extends previous work on the early stage of technology-based entrepreneurship which

is sparse and contradictory. It bases the argumentation on the unclear significance of planning

activities in the context of entrepreneurial activity. The theoretical work on effectuation has produced

a logic that seems a valuable addition to the tools offered in the context of planning and prediction.

Accepting a certain significance of planning or causal reasoning, the basic assumption in this study is

that in the venturing process we can identify both causal and effectual activity. To our knowledge,

there has so far been no study on the phenomenon of effectuation in real-life context. Studies so far

were mainly based on thought experiments or single case histories.

Based on this discussion, we therefore ask the following research questions: Which kind of

behaviour does effectual vs. causal logic induce in venture projects? How do patterns in the

combination of effectual and causal approaches relate with a variation of process performance? How

do effectual and causal behaviour merge into a mechanism for the early-stage process that allows to

cope with uncertainty?

We enhance the research activity field of technology entrepreneurship by looking into the early

phase of technology ventures, adding the concept of effectuation and its differentiation from

causation to the discussion of facing uncertainty. Through in-depth case studies of 8 technology-

based venture projects, we specify the role of uncertainty in the early-stage process, differentiating

between resource and knowledge uncertainty as the predominant types of uncertainty in the early-

stage process and environmental uncertainty as predominant in later stages. We further find that a

variation in gestation process performance (range: 0,05 to 0,27 gestation steps per month over all

cases) matches 4 different patterns of effectual and causal approaches in the early-stage process.

The article begins by providing an overview of the entrepreneurial process literature and the need

for an exploration of uncertainty in this context. We then detail the models of causation and

especially effectuation from entrepreneurial cognition literature in connection with a resource-based

perspective as the theoretical underpinnings to our work. After presenting our research design, we

summarize our data to illustrate and finally discuss our key findings.

31

UNCERTAINTY AND PERFORMANCE IN THE ENTREPRENEURIAL PROCESS

Process

The outcomes of entrepreneurial activity and their benefits have been shown quite often.

However, we still have very limited information on how new businesses come into existence.

Gartner et al. state that “we can see the success of entrepreneurial activity, yet we have few insights

into why particular entrepreneurial efforts were successful while other efforts failed” (Gartner et al.,

2004). They continue that “we have almost no information on the number and characteristics of the

nascent entrepreneurs who attempt to start businesses and the likelihood that such attempts will result

in the formation of new businesses.”

We therefore adopt for our study a process view on technology entrepreneurship in order to

generate more information about how technology entrepreneurs proceed in their process of

technology commercialization. Clarysse and Moray used a similar approach in their study (Clarysse

& Moray, 2004). We seek to explore the same research objective with our qualitative approach to

develop an in-depth understanding of the early-stage entrepreneurial process. A variety of process

models have been offered in the literature so far, amongst some of the more widely accepted of

which are those of Kazanjian, Bhave and Timmons (Bhave, 1994; Kazanjian & Drazin, 1989;

Timmons, 1999). As we are specifically interested in the early-stage process, we draw upon a

process model developed by Gartner et al. which originates from the context of the PSED panel

study (Gartner et al., 2004). This model focuses on the early stage venturing process and will be used

with modifications in setting up our preliminary research framework.

Uncertainty

Uncertainty is one of the main influence factors for new firms and especially technology-based

ventures. As a consequence it is a conceptual cornerstone for most theories of the lentrepreneur

(McMullen & Shepherd, 2006). Historically the theory of uncertainty in the context of

entrepreneurship has been first picked as a central theme by Richard Cantillon in 1752. He defined

an entrepreneur as “someone who engages in exchange for profit; specifically, he or she is someone

who exercises business judgment in the face of uncertainty” (quoted in (Link & Siegel, 2003)).

Knight focussed on the distinction between risk and uncertainty in his fundamental work from

1921. According to him, risk can be calculated beforehand and transferred into direct costs whereas

uncertainty is unspecific and unpredictable, meaning that e.g. certain unique events can not be

foreseen and thus be connected to probabilities within a-priori decision-making processes (M.

Brouwer, 2002; Wu & Knott, 2006). Uncertainty in the Knightian way is the precondition for profit

making by investors/entrepreneurs: “It is true ‘uncertainty’ and not risk which forms the basis of a

32

valid theory of profits and accounts for the divergence between actual and theoretical competition. If

profits could be calculated before the act, or even if there is a mathematical or a-priori type of

probability of success, these risks can be insured and will be changed into costs” (Maria Brouwer,

2000). Brouwer argues that Knight’s understanding of uncertainty describes the uncertainty of the

economy as an entire system and the outcome of personal ventures of investors/entrepreneurs.

With regard to current research on this topic a general differentiation can be made between

research that has so far mainly concentrated on system-level insights and research with a focus on the

individual-level (McMullen & Shepherd, 2006). The main research target of the latter is the

entrepreneur and its individual actions (authors in this context are e.g. Sarasvathy or Shane).

Furthermore, the action-oriented research stream is of specific importance when analyzing why and

how entrepreneurial action is taking and also not taking place. This second one highlights the ability

and willingness to cope with uncertainty, thus dividing those who act entrepreneurially from those

who do not by their different levels of motivation or risk propensity (e.g., (Douglas & Shepherd,

2000; Knight, 1921; Schumpeter, 1934).

Within the research progress on uncertainty theory in the context of entrepreneurship, different

classifications of the topic have been developed that try to include the different research streams

mentioned above. Milliken differentiates three types of uncertainty from a mainly

organizational/system-level point of view: state, effect and response uncertainty namely

unpredictability of future environment, its influence on the organization and the possible responses to

this state (Milliken, 1987). Looking at Milliken’s dimensions, it becomes obvious that this

environment-related definition of uncertainty does not seem to help a lot for early-stage processes

where there does not yet exist much of an environment to worry about.

In comparison to Milliken, Van Gelderen et al. present a broader view, enhancing the uncertainty

definition and proposing a differentiation of uncertainty in six forms structured along the level of

their analysis within the economic theory. According to them uncertainty occurs on the industry, firm

and personal level (Van Gelderen et al., 2003). These dimensions seem to develop the concept of

uncertainty further towards the situation of the technology entrepreneur. In addition to his five basic

forms of uncertainty spread over the three levels of analysis, Van Gelderen et al. introduce a sixth

form of uncertainty, the information/knowledge uncertainty, as a meta-category of uncertainty that

seems to correspond with the dimensions described by Milliken in 1987.

We conclude that the early-stage entrepreneur would not have to worry about industry-level

uncertainty at this early point in time. It is rather firm-level resource uncertainty and the meta-

category of knowledge uncertainty in terms of goal-ambiguity which is important from early on.

Personal level uncertainty, lastly, is incorporated in the effectual model. If the entrepreneur did feel

uncertain about his capabilities, he would never start the process.

33

Process performance

In order to grasp the variation we liked to look at, we had to think about performance.

Performance issues have been and are still discussed intensively in management literature as a whole.

While entrepreneurship as a field of research is undoubtedly dominated by research work on success

factors, the concept of performance at the same time creates complicated issues in the context of

entrepreneurial activity (Ma & Tan, 2006; Murphy et al., 1996). This is due to the emergent

character of the phenomenon, and researchers still discuss the whole spectrum from classic

performance indicators like sales, employee or profit growth to the notion of mere survival.

It is not our aim for this study to create a viable performance measure in terms of venture or even

technology transfer performance. We much rather incorporate the simpler proxy of gestation process

performance in order to match it with findings about behavioural patterns within the process. To

approach the idea of performance in the early-stage venturing process, we therefore draw on the

literature of gestation process steps (Alsos & Kolvereid, 1998; Carter et al., 1996; Diochon et al.,

2005; Gelderen & Frese, 2000; van Gelderen et al., 2006). This stream of literature suggests that it

might help to differentiate between more and less successful early-stage venturing processes by

looking at speed and intensity of a row of gestation steps that have been carried out. We will use data

on gestation process steps in relation to the gestation time as a proxy to judge upon the gestation

process performance of the technology ventures. Gestation steps that are included in this study as a

synthesis from the mentioned literature are: team formation, business planning, grant acquisition,

patent application, prototype presentation, full time engagement, funding acquisition, formal

incorporation, product presentation, facility procurement, employee hiring, sales. We summarize that

we will include a process view, the idea of an initial resource base and a notion of performance in

terms of the gestation process performance in our preliminary research framework.

CONNECTING THE RESOURCE BASE, CAPABILITIES AND THE CONCEPTS OF CAUSATION AND EFFECTUATION

Among others, Alvarez and Busenitz dicuss that the entrepreneurial task involves resource seeking,

accumulation and combination to a large extent, especially in early phases (Alvarez & Busenitz,

2001). Hence, a study of the early-stage entrepreneurial process is likely to also be a study of

resources and resource-related activities. The effectual approach will show to be rather means- or

resource-driven. However, literature on a more advanced and dynamic version of the resource-based

view has shown that it is about what is being done with the resources what counts (Foss & Ishikawa,

2007; Helfat & Peteraf, 2003; Hjorth, 2007). It is therefore interesting to ask how organization theory

might apply to the context of the emergent organization. We see the initial resource base as the

critical base for the venture, whether or not it finally ends up to be creating a competitive advantage

34

for the future organization. However, we feel that there is definitely need for a capability to move

forward. With resource uncertainty being assumed as one of two predominant uncertainty

dimensions, it is of interest how this issue is being solved. Being interested in how the initial

resource base of a venture project is being leveraged and how uncertainty is overcome, we finally

draw upon the concept of dynamic capabilities, trying to find some proof of a dynamic capability that

allows technology entrepreneurs to navigate through the highly ambiguous early stage (Teece et al.,

1997; Zahra et al., 2006).

So far start-up companies have mainly been studied as institutions built around opportunities in

the context with tools like business planning, market research, competitive analysis, acquisition of

resources, etc. (Bygrave, 1989; S. D. Sarasvathy & Dew, 2005b; Shane & Venkataraman, 2000). Yet,

this toolset leaves the entrepreneur with unanswered questions concerning a favourable behaviour

when facing high levels of uncertainty as we find in the context of early-stage technology

entrepreneurship. A couple of recent studies already argue in favour of a more specified and

diversified toolset for entrepreneurs. Ozgen and Baron discovered that specifically three sources of

social opportunity-related information (mentors, informal industry networks and participation in

professional forums) had positive effects on opportunity recognition by entrepreneurs (Ozgen &

Baron, 2007). The study shows the growing interest in a more detailed catalogue of advice for the

early phase of the venturing process. Corbett discusses opportunity recognition in connection with

learning (Corbett, 2007), Gruber propose a more sophisticated view on planning (Gruber, 2007) and

others emphasize the importance of engaging with other parties (Rickne, 2006; Rothaermel & Deeds,

2006); (Morse et al., 2007). Sarasvathy in her work has identified the traditional toolset as being

prediction-based. She labelled it with the term causation to differentiate it from her newly developed

concept of effectuation.

Effectuation tries to solve the problem of uncertainty by introducing an approach for an

opportunity creation process as opposed to opportunity recognition and development. It stems from

the rather young research stream of entrepreneurial cognition, more explicitly from a study on expert

entrepreneurs (S. D. Sarasvathy, 2001), and builds on work of March who proposed a technology of

foolishness to complement the technology of reason (March, 1982). Entrepreneurial cognition again

is based in cognitive psychology which examines internal mental processes such as problem solving,

memory or language. A couple of studies have laid the ground for the nexus of entrepreneurship and

cognition research (Mitchell et al., 2004); (Baron & Ward, 2004); (Gaglio, 2004); (Hindle, 2004). It

is finally Sarasvathy who gives an introductory view on the effectuation logic, claiming that we need

to move to a theory of firm design (S. D. Sarasvathy, 2004).

Sarasvathy developed five principles which form the basis of the effectual logic (S. D.

Sarasvathy, 2001); (S. D. Sarasvathy, 2008). Throughout other studies on effectuation, this number

varied (e.g. (Wiltbank et al., 2006). For our discussion we altered the set of dimensions to four,

35

assuming that the fifth principle distinguishes between the general notion of action- versus

prediction-based, which can be found in any of the four remaining dimensions as the basic reasoning.

Table 1 contrasts effectual and causal logic along four dimensions.

Table 1: Causal versus effectual processes (Source: Based on (S. D. Sarasvathy, 2008)

Avoiding surprises

Prediction of the future

Expected returns

Goal (ends)-driven action

… in causal logic

Dealing with contingencies by …

Moving forward through …

Investment decision based on …

Start of process with …

Process dimensions …

Leveraging contingencies (Lemonade)

Stakeholder commitments (Patchwork quilt)

Commitment limit (Affordable loss)

Means-driven action (Bird-in-hand)

… in effectual logic

Avoiding surprises

Prediction of the future

Expected returns

Goal (ends)-driven action

… in causal logic

Dealing with contingencies by …

Moving forward through …

Investment decision based on …

Start of process with …

Process dimensions …

Leveraging contingencies (Lemonade)

Stakeholder commitments (Patchwork quilt)

Commitment limit (Affordable loss)

Means-driven action (Bird-in-hand)

… in effectual logic

Iterative, converging process that producesa growing set of means and a convergenceon goals, thereby creating a firm

Linear, planned process that deducesthe necessary means by breakingdown predefined goals

A causal process starts with goal-driven action, deducing the necessary resource-base by

breaking down the goals. Investment decision are made on the basis of expected returns, all assuming

that scenarios of future states of the environment can be envisioned. Every step that carries the

project further in the process is based on prediction of the future which logically leads to the

reasoning behind the last dimension that on the predicted way towards the defined goals surprises are

rather to be avoided.

An effectual process on the other hand starts with means-driven action, literally a rigorous

analysis of who I am, what I know and whom I know. With that bird in hand subsequent investment

decision by any party are made on the basis of a commitment limit in terms of an affordable loss. In a

worst case scenario, this investment criterion leads to limited and acceptable loss on the

stakeholder’s side. The direction of development is highly influenced by stakeholders who self-select

themselves to the project through concrete commitments, thereby enhancing the resource-base and

adding their own ideas and goals to the process. The unforeseen, in terms of contingencies, changes

here in character, almost becoming a resource itself, being embraced and leveraged into new options

for the project.

The discussion of the basic principles in terms of process dimensions already hinted at the

dynamic process model that will be applied in our study (S. D. Sarasvathy & Dew, 2005b; Wiltbank

et al., 2006). The shape of the product, the market and eventually the firm comes into existence

through various iterations of this process, incorporating a growing set of means as well as

considering a growing set of goals. The process ends when stakeholder commitments have

converged to precise goals that can no longer be easily changed. From here the company develops

either through switching to causal logic and tools or because the project is being abandoned. Harting

has displayes the dynamics of such a transition in his working paper on Circuit City’s CarMax used

car retailing unit (Harting, 2004).

36

Initial resourcebase:

- Technology entrepreneur

- Technology- Parent

organization- …

Gestation processperformance

UNCERTAINTY

RESOURCES GESTATION PROCESS PERFORMANCE

Causation Effectuation

DYNAMIC CAPABILITY

Figure 1: the preliminary research framework

We conclude our theoretical discussion by adding the concepts of causation and effectuation to

our preliminary research framework in Figure 1, proposing that it is a valuable research endeavour to

analyze case study material in order to learn more about the concepts’ significance within the early-

stage technology entrepreneurship process.

METHODOLOGY AND EMPIRICAL DESIGN

A couple of studies have looked into what is called the nascent stage of entrepreneurial activity

so far (Diochon et al., 2005, 2007; Liao et al., 2005; Wood & Brown, 1998). However, in general the

field is characterized by difficulties in access to emerging projects and data collection within the field

so that not many studies have focussed on the venturing process in very early stages. By using a

multi-case case study approach, we examine the early process steps in the context of technology-

based entrepreneurship and the creation of new technology-based firms. We have chosen 8 cases that

are based on technologies, and we focus on the venture development process. The projects in this

study’s sample are originating from Germany’s two leading technological universities, Aachen and

Munich. In order to enhance validity of results, we first gathered data at RWTH Aachen University

and then replicated in two out of three groups one case each at the Technical University of Munich in

order to enhance validity.

Choosing a case study approach

The choice for a qualitative approach was made for two reasons. First, the nature of the research

question requires the collection of in-depth qualitative data. Effectuation has been detected in expert

entrepreneurs as a cognitive process. Hence, examining to what extent and how the development

process of a technology-based venture is characterized by effectual elements requires a cumulative

and partially iterative process of information gathering from parties that influenced the period of

early venturing activity. Yin and Eisenhardt both agree that one ore more cases can be used and that

37

the data should usually come from multiple data sources (K. M. G. Eisenhardt, Melissa E., 2007;

Yin, 2003). Experimental settings do not allow for information about the current level of effectual

behaviour in the context of real venture projects. Survey-based settings only catch very limited and

pre-structured information.

Second, research on the early process of technology-based entrepreneurship has been sparse and

contradictory. Research in the field of effectuation has been conducted since 1998 (D. K. Sarasvathy

et al., 1998). Therefore we are coping with a rather underdeveloped field of research which,

however, gains more and more attention. Recognizing this stage of development, the study employs

an explorative approach towards the research questions (Yin, 2003). This loosens the restrictions on

data gathering and allows, in turn, for a more multi-faceted view on reality. We also refer to

Edmondson and McManus who claim that qualitative research answers how and why questions in

comparably unexplored areas (Edmondson & McManus, 2007). Eisenhardt defines case study

research as an empirical research strategy with the goal to inductively build or develop theory (K. M.

Eisenhardt, 1989). With regard to theory, researchers have yet to converge on clear-cut constructs in

order to operationalize effectuation for large-scale testing. The case study method might therefore

also allow for a more comprehensive description of the notions of causation and effectuation in order

to further develop the theory.

Sample selection

We have aimed at purposeful sampling of cases to the extent possible (Neergaard, 2006), (Patton,

2002). The total list of cases we looked at came from a project list of the center for entrepreneurship

at RWTH Aachen University. We explicitly excluded all companies that were not spin-offs from

Aachen University in the first place. The preliminary list of German spin-off projects from RWTH

Aachen University contained 59 projects. Out of these 59 projects we selected cases by applying the

following selection criteria: first of all they had to be technology-based. However, we excluded any

projects with a pure biotech, pharma or medical device focus for their specificities in terms of

gestation process. With regard to the age of the projects, we agreed to look for projects with gestation

activities having started at least 3 years ago and incorporation not having occurred more than 3 years

ago. Through these criteria we wanted to make sure that internal and external stakeholders could still

remember the nascent stage but would also be able to comment on a reasonable gestation phase. We

ended up with 8 projects, eliminating most projects for a too short gestation phase or unavailability

for an in-depth study. After a first round of interviews, we eliminated another 2 projects for a strong

market-pull basis or too complex technology development process, leaving us with 6 cases to

present. In the replication processes with started out with 3 cases and eliminated one due to a market-

pull basis again. We will therefore present data on 8 cases in total.

38

Data and measures

For each case, information was gathered in interviews and from company documents. We

structured the interview guideline along our preliminary research framework, inquiring for the initial

resource base (technology, technology’s organizational background, personal background and

involvement), the early-stage process and the current development stage. Interviews were carried out

with three to four stakeholders, both internal and external. Further data was gathered from the

internet and material that the entrepreneurs provided. The data was transcribed and coded,

differentiating the semantic chunks as indicators for either causal or effectual behaviour in one of the

4 process categories. The coding allowed us to create the measure of effectuation rate (effectuation-

related semantic chunks over the sum of causal and effectual semantic chunks) as a proxy measure

for effectuation intensity. As a second important measure we used gestation speed as a proxy for

success, calculating it as the quotient of the number of gestation steps over the number of gestation

months. Results were evaluated within a case-based as well as a cross-case analysis.

We have taken a number of measures to try to ensure as much objectivity as possible in the data

collection and interpretation process. Theses activities strive for a triangulation of data, in our case

mainly through multiple data sources (interviews and documents). For coding reliability, we

appointed a second coder and discussed all divergent classifications. Finally, we are aware of further

concerns with regard to the case method, e.g. retrospective bias and self-justifying explanations in

connection with interviewees which we also try to address through triangulation of both internal and

external stakeholders.

CASES VARY IN THEIR APPLICATION OF CAUSATION AND EFFECTUATION TO COPE WITH UNCERTAINTY

In the following two sections we turn to present and discuss our analysis and results. Causal

mechanisms and tools have been predominantly studied in the context of early-stage entrepreneurial

activity and technology entrepreneurship. However, current research still comes up with new insights

into these toolsets and keeps adding notions of networking, personal resources and the role of

uncertainty. Sarasvathy, coming from a different angle, finally developed the compelling process of

effectuation that seems to fit our research object (S. D. Sarasvathy, 2008). This discussion and our

own experiences about variation among technology ventures in terms of gestation process

performance led us to carry out this study.

We found that the variation in terms of gestation speed can be explained through a different level

of effectuation intensity in the early-stage process (correlation of 0,62). Our data revealed three

different groups that are displayed in Table 2. We have disguised the company names, referring to

39

them through names of Saturn moons with Calypso and Narvi being the replicated cases at the

Technical University of Munich.

Table 2: Summary of case studies

1 / 2

1 / 2

2 / 0

1 / 2

1 / 2

1 / 3

2 / 2

2 / 1

Number of interviews(founders/

other)

64%0,12668Quality inspectiontool

Atlas

„low speed“Pattern: rather causal approach

Bestla: comparably high level of effectuation

39%0,09323Long-term electricityprovider

Narvi

40

46

54

51

45

41

Gestationmonths

(until12/07)

„moderate speed“Pattern: effectual outset but no clear switch to causal activities

Greip: rather causal approachfrom beginning

„high speed“Pattern: switch from effectual to causal activities

Group description

Bestla

Mimas

Greip

Pan

Calypso

Methone

Venture project*

High-energy pellets

Biotech chips

Special surfaceproducts

Laser components

Wafer handlingmachines

Biotech fermentationtool

(Potential) Product

2

4

9

10

12

11

Gestationsteps

0,05

0,09

0,17

0,20

0,27

0,27

GestationSpeed(steps/ month)

57%

46%

50%

74%

70%

63%

EffectuationRate

(percentageof effectualstatements)

1 / 2

1 / 2

2 / 0

1 / 2

1 / 2

1 / 3

2 / 2

2 / 1

Number of interviews(founders/

other)

64%0,12668Quality inspectiontool

Atlas

„low speed“Pattern: rather causal approach

Bestla: comparably high level of effectuation

39%0,09323Long-term electricityprovider

Narvi

40

46

54

51

45

41

Gestationmonths

(until12/07)

„moderate speed“Pattern: effectual outset but no clear switch to causal activities

Greip: rather causal approachfrom beginning

„high speed“Pattern: switch from effectual to causal activities

Group description

Bestla

Mimas

Greip

Pan

Calypso

Methone

Venture project*

High-energy pellets

Biotech chips

Special surfaceproducts

Laser components

Wafer handlingmachines

Biotech fermentationtool

(Potential) Product

2

4

9

10

12

11

Gestationsteps

0,05

0,09

0,17

0,20

0,27

0,27

GestationSpeed(steps/ month)

57%

46%

50%

74%

70%

63%

EffectuationRate

(percentageof effectualstatements)

Correlation (gestation speed / effectuation rate): 0,62 (without Bestla: 0,71)

In group 1 we see projects that not only showed effectual elements in their early stage. They also

seemed to be able to transition to a more causal approach once a resource base was established and a

development path for the project definable. However sophisticated their current development seems

to be, they still started with very basic questions and answers rooted in the bird-in-hand principle:

“And about me – I really did not feel like going to a large company. And then the question came

up, whether I want in some small company, well, some responsibility there, or actually doing it

myself. And so we said to ourselves: we got this option to use this technology – let’s try and see

how we can fund that … So, from that point of view, I always enjoyed to lead a team, to lead a

team and work conceptually. And I did that in many areas. I also led a project management

group once, as a tutor. And, yes, creative work is just exciting. And where can I work better

creatively than either as head of a development department or right away in my own company.”

(Quote from an interview with Calypso)

Group 3 is easy to define as well. We found only very few or no effectual elements in these

processes showing that the projects relied more on the causal approach. We conclude that these

projects could need training in effectuation to move forward. Up to now, these cases would probably

have been discarded with the explanation that the individuals are not entrepreneurial enough. Of

course, we are not able to evaluate whether teaching effectuation can fundamentally change a believe

in causal mechanisms. This quote shows some confusion in terms of causal concepts:

40

“The plan was to put the whole thing to reality as quickly as possible. That might have been a

little too ambitious. Then of course the search for funding … is actually much more difficult than

I would have expected. And then you also need a team, that wants to be built – which again is

quite a dilemma. Without money you can hardly build a team, or only with difficulties. And

without a team you don’t get money. There the cat bites its own tail.” (Quote from interview with

Mimas).

Group 2 is a little more difficult to grasp. The projects either show quite a lot of effectual

elements within the process but seem to have trouble to transition to more causal approaches. The

project Greip forms its own subgroup on the other hand due to an interesting reliance on causal

elements. The project took a long time to develop but achieved a reasonable process performance in

terms of building up a company.

In many of the more effectual cases, it can be shown that personal preferences, knowledge and

networks are important sources for early venturing activities. This usually also holds for the cases

that then relied more intensely on causal approaches. We were also able to trace the process of

stakeholder acquisition. Many entrepreneurs sought exposure to a variety of possible stakeholders,

allowing for the self-selection process to take place.

Interestingly enough, in group 1 and even partly in 2, even when a formal business plan had been

written, this tool did not guide the venture creation process but served as a communication tool

targeted towards the acquisition of stakeholders that did not belong to the entrepreneur’s peer group.

What is more, many projects have successfully created products and did so under the influence of

stakeholders.

OVERCOMING UNCERTAINTY BY COMBINING EFFECTUAL AND CAUSAL APPROACHES

Enhancing the toolbox of early-stage technology entrepreneurs (and maybe even entrepreneurs in

general) through elements of effectuation seems to pay off. Cases that were able to start out

effectually and then transition to more causal approaches went quickly through their gestation

process. Other projects that relied more intensely on causal concepts took much longer or actually

got stuck in their process, which might hint at a necessity to apply more effectual elements in their

process. With regard to our specific research questions, we can now enhance the discussion of the

effectuation/causation toolboxes as follows.

In a first step, we deepened the understanding of the concepts of effectuation and causation. This

was realized through a category-based analysis of the codings and a respective refinement of the

coding scheme as part of the coding process and the subsequent discussion between first and second

coder. It turned out that the four coding categories (Process start, Investment decision basis, Moving

41

forward in the process and Dealing with contingencies – see Table 1) varied in their importance in

terms of the number of respective semantic chunks. Process start and Moving forward were by far

and over all cases the predominant dimensions. Looking more closely, the categories also differed as

to their character: either logic or behaviour. Moving towards operationalization, this implies that

focus should be on Process start and Moving forward in the process, while the logics of investment

ideas and contingency approaches are tied to the core categories. Lastly, it was possible in this step to

depict the notion of predictability as the relevant connector between causation and effectuation. This

three-step analysis (category importance, category character and connector) will facilitate further

work on operationalization in future research.

In a second step, we found that patterns in the combination of effectual and causal approaches

relate with a variation of process performance. Three cases (Methone, Calypso and Pan) showed a

comparably quick movement through the venture process in connection with a high level of

effectuation intensity. Narvi, Mimas and Bestla, on the other hand, showed a comparably slow

venture process in combination with lower levels of effectuation intensity. These results suggest that

effectual behaviour in the early-stage process fosters venture development while causal behaviour is

not as effective in the early stage.

In a last step, we found that effectual and causal behaviour merge into a mechanism that could be

labelled uncertainty management, as both toolboxes seem to address different kinds of uncertainties

in the early stages of the venture development. While effectuation helps the projects in the early

phase to create a resource base and choose a development path without relying on prediction, it is the

prediction-based causal toolset which later helps to define both the environment (e.g. market) for the

venture as well as its prospective position within this environment. It was interesting to see that most

projects showed effectual behaviour especially in the pre-firm phase and switched to causal

behaviour after building a formal company. Especially in the case of Calypso, it was interesting to

see that most effectual behaviour took place at a very early stage, when the staff was still part of the

parent organization, using those resources there to create the resource base for the future venture.

Our data show that it might be worthwhile to further explore the connection between types of

uncertainty and causal and effectual approaches, respectively. While there is not better or worse in

terms of causal or effectual approaches, we grew more confident that causation and effectuation

answer to different types of uncertainty: Effectuation rather addresses early-stage uncertainty in

terms of resource and knowledge uncertainty while causation might rather solve later-stage

uncertainty in terms of environmental uncertainty.

With regard to our theoretical basis, we also see implications in terms of the dynamic resource-

based view and the concept of dynamic capabilities. Given the fact that especially technology

entrepreneurs need to cope with uncertainty in their early venturing process, we would like to

stimulate the discussion about a dynamic capability of uncertainty management for the emergent

42

organization that combines uncertainty identification and distinction with the appropriate application

and combination of both the causal and the effectual toolset.

An important practical implication hints at the set-up of teaching and support programs for

entrepreneurs in general and more specifically technology entrepreneurs. Introducing effectuation to

technology entrepreneurs might help to move from a more “accidental” employment of effectual

logic to a more structured use in start-up processes. If it can be shown to be a reliable success factor,

there is a good chance that explicitly conceptualizing the phenomenon and teaching it can foster

technology transfer within technology ventures. As much, the results could allow for further

empirical insight and feed back to the concept of effectuation and enhance the theoretical

understanding of it.

Limitations

There are of course certain limitations to this study. The most obvious concern is that of

statistical generalizability. A case study, or even eight, can never hold to prove relationships between

constructs. However, it is the goal to enhance theoretical knowledge through case study research, not

to prove theory. A variety of influencing factors can be assumed for any case presented. Although we

tried to look at equally complex technologies, the variation among different technological fields

hinting at an even broader and diverse set of possible markets is definitely a strong influence. We

address this issue by arguing that the high level of uncertainty and the early stage allow for a

comparison between the cases. As much as current literature mainly refers to technology that is being

created at university, we also looked exclusively at spin-offs from university. We argue that

universities are both a strong technology source and not as far developed in terms of transfer

mechanisms which led us to hope that we would be able to find more variation among cases in terms

of the gestation process as would have been possible in more professional environments. Autio even

labels the university spin-off as the “classical case” in terms of spin-off activity (Autio, 1997; Autio

& Yli-Renko, 1998).

Future research

The research opens up some avenues of further research. First of all, in terms of the technologies’

organizational background it is probably worthwhile looking at projects coming out of non-

university public research institutions as well as from the corporate world. Secondly, with regard to

methodology, it would be interesting to look at experimental set-ups. Founders could be taught the

toolset of effectual logic, finally being compared to a control group with founders that did not go

through this kind of teaching. What is more, next steps could aim at statistical generalization for

effectuation as a success driver in new technology-based firms, maybe broadening these endeavours

43

in terms of an a more detailed conceptualization or even operationalization of uncertainty

management, including both causal and effectual approaches.

CONCLUSIONS

The findings in this study are important because they shed light on the early stage of technology

entrepreneurship. Most importantly, we show that venture projects differ in their way of applying

both causal and effectual approaches in their early-stage development process. Our data propose that

those projects that add effectual elements to their toolbox are able to overcome resource and

knowledge uncertainty through expanding means and converging on goals at the same time.

However, at some point the projects have to turn to causal mechanisms in order to build the

organization.

REFERENCES

Alsos, G. A., & Kolvereid, L. (1998). The business gestation process of novice, serial, and parallel

business founders. Entrepreneurship: Theory & Practice, 22(4), 101. Alvarez, S. A., & Busenitz, L. W. (2001). The entrepreneurship of resource-based theory. Journal of

Management, 27(6), 755. Autio, E. (1997). `atomistic' and `systemic' approaches to research on new, technology-based firms:

A literature. Small Business Economics, 9(3), 195. Autio, E., & Yli-Renko, H. (1998). New, technology-based firms as agents of technological

rejuvenation. Entrepreneurship & Regional Development, 10(1), 71. Baron, R. A., & Ward, T. B. (2004). Expanding entrepreneurial cognition's toolbox: Potential

contributions from the field of cognitive science. Entrepreneurship: Theory & Practice, 28(6), 553.

Bell, M., & Pavitt, K. (1993). Technological accumulation and industrial growth: Contrasts between developed and developing countries. Industrial & Corporate Change, 2(2), 157.

Bhave, M. P. (1994). A process model of entrepreneurial venture creation. Journal of Business Venturing, 9(3), 223.

Brouwer, M. (2000). Entrepreneurship and uncertainty: Innovation and competition among the many. Small Business Economics, 15(2), 149.

Brouwer, M. (2002). Weber, schumpeter and knight on entrepreneurship and economic development. Journal of Evolutionary Economics, 12, 83-105.

Bullinger, H.-J. (2002). Technologiemanagement - forschen und arbeiten in einer vernetzten welt. Berlin Heidelberg: Springer-Verlag.

Bygrave, W. D. (1989). The entrepreneurship paradigm (ii): Chaos and catastrophes among quantum jumps? Entrepreneurship Theory and Practice, Winter 1989, 7-30.

Bygrave, W. D., & Hofer, C. W. (1991). Theorizing about entrepreneurship. Entrepreneurship Theory and Practice, Vol. 15), 13-22.

Carter, N. M., Gartner, W. B., & Reynolds, P. D. (1996). Exploring start-up event sequences. Journal of Business Venturing, 11(3), 151.

Clarysse, B., & Moray, N. (2004). A process study of entrepreneurial team formation: The case of a research-based spin-off. Journal of Business Venturing, 19(1), 55-79.

Corbett, A. C. (2007). Learning asymmetries and the discovery of entrepreneurial opportunities. Journal of Business Venturing, 22(1), 97.

Diochon, M., Menzies, T. V., & Gasse, Y. (2005). Exploring the relationship between start-up activities and new venture emergence: A longitudinal study of canadian nascent entrepreneurs. International Journal of Management & Enterprise Development, 2(3/4), 1.

44

Diochon, M., Menzies, T. V., & Gasse, Y. (2007). Attributions and success in new venture creation among canadian nascent entrepreneurs. Journal of Small Business & Entrepreneurship, 20(4), 335.

Douglas, E., & Shepherd, D. A. (2000). Entrepreneurship as a utility maximizing response. Journal of Business Venturing, 15, 393-410.

Edmondson, A. C., & McManus, S. E. (2007). Methodological fit in management field research. Academy of Management Review, 32(4), 1155.

Eisenhardt, K. M. (1989). Building theories from case study research. Academy of Management Review, 14(4), 532.

Eisenhardt, K. M. G., Melissa E. (2007). Theory building from cases: Opportunities and challenges. Academy of Management Journal, 50(1), 25-32.

Foss, N. J., & Ishikawa, I. (2007). Towards a dynamic resource-based view: Insights from austrian capital and entrepreneurship theory. Organization Studies (01708406), 28(5), 749.

Gaglio, C. M. (2004). The role of mental simulations and counterfactual thinking in the opportunity identification process. Entrepreneurship: Theory & Practice, 28(6), 533.

Gartner, W. B., Shaver, K. G., Carter, N. M., & Reynolds, P. D. (2004). Handbook of entrepreneurial dynamics - the process of business creation. Thousand Oaks, California: Sage Publications.

Gelderen, M. v., & Frese, M. (2000). Strategies, uncertainty and performance of small business startups. Small Business Economics, 15(3), 165.

Gruber, M. (2007). Uncovering the value of planning in new venture creation: A process and contingency perspective. Journal of Business Venturing, 22(6), 782.

Harting, T. (2004). Entrepreneurial effectuation in a corporate setting: The case of circuit city’s carmax unit, BCERC Babson College Entrepreneurship Research Conference. Strathclyde, Scotland: Babson College.

Helfat, C. E., & Peteraf, M. A. (2003). The dynamic resource-based view: Capability lifecycles. Strategic Management Journal, 24(10), 997.

Hindle, K. (2004). Choosing qualitative methods for entrepreneurial cognition research: A canonical development approach. Entrepreneurship: Theory & Practice, 28(6), 575.

Hjorth, D. (2007). Lessons from iago: Narrating the event of entrepreneurship. Journal of Business Venturing, 22(5), 712.

Hsu, D. H., Roberts, E. B., & Eesley, C. E. (2007). Entrepreneurs from technology-based universities: Evidence from mit. Research Policy, 36(5), 768.

Kazanjian, R. K., & Drazin, R. (1989). An empirical test of a stage of growth progression model. Management Science, 35(12), 1489.

Kirchhoff, B. A., Newbert, S. L., Hasan, I., & Armington, C. (2007). The influence of university r & d expenditures on new business formations and employment growth. Entrepreneurship: Theory & Practice, 31(4), 543.

Klofsten, M. (2005). New venture ideas: An analysis of their origin and early development. Technology Analysis & Strategic Management, 17(1), 105-119.

Knight, F. H. (1921). Risk, uncertainty and profit. New York: Houghton Mifflin. Liao, J., Welsch, H., & Tan, W.-L. (2005). Venture gestation paths of nascent entrepreneurs:

Exploring the temporal patterns. Journal of High Technology Management Research, 16(1), 1.

Licht, G., & Nerlinger, E. (1998). New technology-based firms in germany: A survey of the. Research Policy, 26(9), 1005.

Link, A. N., & Siegel, D. S. (2003). Technological change and economic performance. New York: Routledge.

Ma, H., & Tan, J. (2006). Key components and implications of entrepreneurship: A 4-p framework. Journal of Business Venturing, 21(5), 704.

March, J. G. (1982). The technology of foolishness. In J. G. March & J. P. Olson (Eds.), Ambiguity and choice in organizations (pp. 69-81). Bergen.

McMullen, J. S., & Shepherd, D. A. (2006). Entrepreneurial action and the role of uncertainty in the theory of the entrepreneur. Academy of Management Review, 31(1), 132.

Mellewigt, T., & Witt, P. (2002). Die bedeutung des vorgründungsprozesses für die evolution von unternehmen: Stand der empirischen forschung. Zeitschrift für Betriebswirtschaft, 72. Jg.(Heft 1), 81-109.

Milliken, F. J. (1987). Three types of perceived uncertainty about the environment: State, effect and response uncertainty. Academy of Management Review, 12(1), 133-143.

Mitchell, R. K., Busenitz, L., Lant, T., McDougall, P. P., Morse, E. A., & Smith, J. B. (2004). The distinctive and inclusive domain of entrepreneurial cognition research. Entrepreneurship: Theory & Practice, 28(6), 505.

45

Morse, E. A., Fowler, S. W., & Lawrence, T. B. (2007). The impact of virtual embeddedness on new venture survival: Overcoming the liabilities of newness. Entrepreneurship: Theory & Practice, 31(2), 139.

Murphy, G. B., Trailer, J. W., & Hill, R. C. (1996). Measuring performance in entrepreneurship research. Journal of Business Research, 36(1), 15.

Neergaard, H. (2006). Sampling in entrepreneurial settings. In H. Neergaard & J. P. Ulhoi (Eds.), Handbook of qualitative research methods in entrepreneurship (pp. 253-278). Cheltenham: Edward Elgar.

Niefert, M., Metzger, G., Heger, D., & Licht, G. (2006). Hightech-gründungen in deutschland: Trends und entwicklungsperspektiven. Mannheim: Zentrum für Europäische Wirtschaftsforschung (ZEW).

Ozgen, E., & Baron, R. A. (2007). Social sources of information in opportunity recognition: Effects of mentors, industry networks, and professional forums. Journal of Business Venturing, 22(2), 174.

Patton, M. Q. (2002). Qualitative research & evaluation methods (3 ed.). London: SAGE Publications.

Reynolds, P. D., & Miller, B. (1992). New firm gestation: Conception, birth and implications for research. Journal of Business Venturing, 7(5), 405-417.

Rickne, A. (2006). Connectivity and performance of science-based firms. Small Business Economics, 26(4), 393.

Rothaermel, F. T., & Deeds, D. L. (2006). Alliance type, alliance experience and alliance management capability in high-technology ventures. Journal of Business Venturing, 21(4), 429.

Sarasvathy, D. K., Simon, H. A., & Lave, L. (1998). Perceiving and managing business risks: Differences between entrepreneurs and bankers. Journal of Economic Behavior & Organization, 33(Issue 2), 207.

Sarasvathy, S. D. (2001). Causation and effectuation: Toward a theoretical shift from economic inevitability to entrepreneurial contingency. Academy of Management Review, 26(2), 243-264.

Sarasvathy, S. D. (2004). Making it happen: Beyond theories of the firm to theories of firm design. Entrepreneurship: Theory & Practice, 28(6), 519.

Sarasvathy, S. D. (2008). Effectuation: Elements of entrepreneurial expertise. Cheltenham: Edward Elgar.

Sarasvathy, S. D., & Dew, N. (2005a). Entrepreneurial logics for a technology of foolishness. Scandinavian Journal of Management, 21(Issue 4), 385-406.

Sarasvathy, S. D., & Dew, N. (2005b). New market creation through transformation. Journal of Evolutionary Economics, 15(Issue 5), 533-565.

Schumpeter, J. A. (1934). The theory of economic development. New Brunswick, NJ: Transaction. Shane, S. (2004a). Academic entrepreneurship: University spin-offs and wealth creation.

Cheltenham: Edward Elgar. Shane, S. (2004b). Encouraging university entrepreneurship? The effect of the bayh-dole act on

university patenting in the united states. Journal of Business Venturing, 19(1), 127-151. Shane, S., & Venkataraman, S. (2000). The promise of entrepreneurship as a field of research.

Academy of Management Review, Vol. 25(Nr. 1), 217-226. Teece, D. J., Pisano, G., & Shuen, A. (1997). Dynamic capabilities and strategic management.

Strategic Management Journal, 18(7), 509. Timmons, J. A. (1999). New venture creation - entrepreneurship for the 21st century (5. Auflage

ed.). Boston u. a.? Van Gelderen, M., Frese, M., & Thurik, A. R. (2003). Strategies, uncertainty and performance of

small business startups. Unpublished Paper. van Gelderen, M., Thurik, R., & Bosma, N. (2006). Success and risk factors in the pre-startup phase.

Small Business Economics, 26(4), 319. Wiltbank, R., Dew, N., Read, S., & Sarasvathy, S. D. (2006). What to do next? The case for non-

predictive strategy. Strategic Management Journal, 27(10), 981. Wood, S. C., & Brown, G. S. (1998). Commercializing nascent technology: The case of laser diodes

at sony. Journal of Product Innovation Management, 15(2), 167. Wu, B., & Knott, A. M. (2006). Entrepreneurial risk and market entry. Management Science, 52(9),

1315. Yin, R. K. (2003). Case study research: Design and methods (3rd ed.). Thousand Oaks et al.: Sage

Publications. Zahra, S. A., Sapienza, H. J., & Davidsson, P. (2006). Entrepreneurship and dynamic capabilities: A

review, model and research agenda. Journal of Management Studies, 43(4), 917.

46

47

FUNDING NTBFS: BEYOND VENTURE CAPITAL, THE ROLE OF DEBT

Tommaso Minola Department of Management, Economics and Industrial Engineering, Politecnico di Milano, Italy

Giuseppe Colombo, 40–20133 Milan, Italy. Tel.: +39 02 23992992; fax: +39 02 23992992

tommaso.minola@polimi.it (corresponding author)

Marco Giorgino Department of Management, Economics and Industrial Engineering, Politecnico di Milano, Italy

Giuseppe Colombo, 40–20133 Milan, Italy. marco.giorgino@polimi.it

ABSTRACT

The objective of our research is to understand which factors at the firm level influence access to external capital for new technology based firms (NTBFs), a puzzle that the extensive literature has not solved. We specify a multi-level framework for the empirical analysis of determinants of NTBF financing performed on a UK dataset. The original contribution of our research is the use of an expert system (Bayesian network) to shape the model of the NTBF fundraising process. Our research results in weak acceptance of pecking order theory (POT) and identifies some qualifying accompanying factors, variables at firm level that have a clear and significant role in the fundraising process. Contrary to the mainstream view, banks can play a role in fostering technology-based entrepreneurship, especially in advanced contexts such as the UK, where the availability of public guarantee schemes has proved effective.

Keywords: Pecking order theory, debt financing, NTBFs, Bayesian network

48/101

INTRODUCTION

The generation and exploitation of knowledge play a large part in creating economic value

(Gibbons et al 1994) in a globalised world; knowledge, human capital and technology development

are central to the growth and wealth of nations (Harris 2001).

Small and medium-sized enterprises (SMEs) and new technology based firms (NTBFs) – the

main players in the knowledge economy – have become the focus of investigation by scholars and

policy makers, because, in OECD economies in particular, ‘radical changes in ICT and

biotechnology have created market opportunities that are more effectively developed by new firms

than by established companies. The shift to knowledge-based economic activity is said to be the

driving force underlying the emergence of the entrepreneurial economy’ (Stam and Garnsey 2007).

This emphasis on the entrepreneurial economy has generated much of policy makers’ and

academics’ interest in NTBFs. We adopt Rickne and Jacobsson’s (1999) definition of an NTBF as

‘a firm whose strength and competitive edge derives from the know-how within natural science,

engineering or medicine of the people who are integral to the firm, and upon the subsequent

transformation of this know-how into products or services for a market’ (Rickne and Jacobsson

1999). NTBFs operate in innovative and technology intensive industries, such as electronic

engineering, computer science, engineering physics, industrial economics, chemical engineering,

mechanical engineering, civil engineering and medicine (Rickne and Jacobsson, 1999). For some

authors, ‘new’ means firms that are no older than 25 years; for the purpose of our research, since

our interest is in the financial attitude of firms at start-up, we focus on early stage companies, in the

very first years of their development.

Before they can show significant growth and success, one of the most significant constraints

new firms have to face is the lack of resources. It seems that NTBFs face a particular financial gap

since the combination of high-tech risk, information asymmetries and low collateral contribute to

funding gaps (Carpenter and Petersen, 2002).

We are interested in external sources of capital for NTBFs, borrowing from Berger and Udell’s

(1998) general scheme of external financial sources for SMEs. According to Berger and Udell, the

modern information-based theory of security design predicts that the mix of external equity and

debt sources will be affected by three dimensions of informational opacity/costly state verification,

adverse selection (tending to favour debt), and moral hazard (tending to favour equity). Differences

in the relative severity of these information problems may help explain how SMEs obtain money.

Cosh et al (2005) found that completely new start-ups are not likely to attract finance from

banks, and equity financing is often said to be the best solution for the financing of young high-tech

companies – because of the high uncertainty of returns, the low probability of financial success,

limited collateral and substantial information asymmetries between firms and investors (Carpenter

49/101

and Petersen 2002). However, there are NTBFs that do receive debt capital but not equity. In

previous research, we studied those firms and tried to understand which factors enable this

phenomenon. In-field pilot interviews allowed us to focus on the main aspects of the problem

(Minola and Giorgino 2008). For the literature, we specified a multi-level framework of analysis

that was used, refined and preliminarily validated on UK and Italian cases (Minola et al 2007). The

object of our present research is an empirical analysis of the determinant of NTBF financing using

this framework and performed on a dataset on UK NTBFs.

One of the original contributions of our research is the use of an expert system to shape the

model of the NTBF fundraising process. An expert system is a software system that incorporates

concepts derived from experts in a field and uses their knowledge to provide problem analysis

through rules. An expert system may also provide statistical analysis of the problem and have the

great advantage of being able to handle complex phenomena, where the definition of variables and

relationships is mainly uncertain and vague.

Although there is evidence that the pattern of lending to young and technology based businesses

can be inverted by innovative bank-lending programs (as described below), bank financing may be

inadequate to fund NTBFs – if it is restricted to the traditional banks. A mismatch of expectations

between banks and technology firms – often referred to as ‘the empathy gap’ (Gill et al 2007) – has

proved a persistent barrier to greater engagement by banks in the tech sector.

Nevertheless banks remain the most common form of external financing for small businesses

and, despite theoretical issues and general declaration, do lend to NTBFs, under certain

circumstances. Fluck et al (1997) stated that information opacity does not prevent firm’s attempts to

obtain external financing, particularly debt to start-up firms, and Berger and Udell (1998) found that

debt appears to be an important source for young firms, contrary to conventional wisdom, thanks to

substantial recourse to personal wealth and collateral.

Banks play a central role in the development of SMEs in general, since almost 8 out of 10 firms

reportedly use bank debt, while only 11 per cent of them receive capital from public institutions. An

even smaller percentage are financed by private investors (seven per cent) and venture capital

companies (two per cent) (Flash Eurobarometer 2005).

In the UK, banks remain the most common form of external financing for small businesses

(according to the British Bankers’ Association, BBA), the main UK retail banks have 3.1 million

small business customers with under £1m turnover). There are four main reasons for this:

familiarity with bank managers; easy access to information; the cost of financing (only the fastest-

growing companies will be able to access large sums of venture capital or other equity finance); and

flexibility (see Pavlov et al 2004).

50/101

UK banks also work with organisations such as the Prince’s Trust, which guarantees small

business lending, and the government to fund the Small Firms Loan Guarantee (SFLG) scheme. The

SFLG covers 75 per cent of the lenders’ exposure, with the borrower paying a two per cent

premium to the government. The aim of the SFLG is to help viable, debt-appropriate businesses that

lack sufficient collateral to access loan financing in the market.

The object of our present research is to understand which factors, at the firm level (i.e.

technology, financial constraints, assets composition, market, and size of business) influence the

access to external capital, in particular bank loans and venture capital financing. Our research

question is: when is debt viable for NTBFs? And furthermore, which factors influence the choice of

the source of external financing for an NTBF?

The issue is relevant primarily for scholars, since financial theory creates a puzzle: ‘Some

authors argue that high-risk entrepreneurs choose debt contracts instead of equity contracts since

risky but high returns are of relatively more value for a loan-financed firm’ (Schäfer et al 2004).

Contrary to this view, entrepreneurial finance theory commonly states that the more risky their

projects, the more likely entrepreneurs are to seek equity contracts. Our work aims to take a first

step toward resolving this puzzle, providing an empirical specification of the variables that

influence choice at firm level.

Our research is also relevant to entrepreneurs: we attempt to develop a framework to help

technology-based entrepreneurs match their business plans with the most appropriate financial

strategy. As we have already said, the literature, although vast, has not solved the puzzle of

entrepreneurial financing; yet, from an empirical point of view, banks are moving toward

innovative offers, trying to meet the financial needs of start-up, and addressing situations where

entrepreneurs are unlikely to attract VC investment but whose projects are nonetheless worth

financing. On the other hand, spreading awareness of the financial strategies of NTBFs may support

entrepreneurs who are VC-worthy but unaware of it.

Finally, our research also concerns policy makers. NTBFs are an important focus for those who

deal with competitiveness policies, as they give significant propulsion to economic growth, in terms

of employment (Acs 2004), and are a primary source of radical innovation (Audretsch 1995).

Governments should be concerned about the possibility of (eventual) market failure preventing

NTBFs from accessing the financial resources they need – although as far as the existence of a real

gap in financial markets is concerned, the literature shows contrasting patterns (see Cosh et al 2005;

De Meza and Webb 1987).

In Section 2 we describe empirical evidence of NTBF financing and review related literature

streams. Section 3 and 4 describe the data and the methodology we adopted. Section 5 develops the

51/101

empirical analysis, and the results are analysed in Section 6. Section 7 summarizes and concludes

the paper.

LITERATURE REVIEW AND HYPOTHESIS FORMULATION

The acknowledged starting point for all entrepreneurial finance theories is irrelevance theory

(Modigliani and Miller, 1958), according to which, in a perfect capital market – that is, one without

taxes, information asymmetries and transaction costs – the market value of a company does not

depend on its financial gearing. The interesting point of this approach is that it shows that, because

capital structure does indeed matter, one or more previous assumptions have been violated. In fact,

irrelevance theory generated concern in modern economical theory about information asymmetries

and principal-agent relationships (Rothschild and Stiglitz 1976 and Holmstrom and Tirole 1997),

maintaining that entrepreneurs always have more complete information about their ventures than

the investors to whom they turn for support. These asymmetries can become a serious issue and

result in moral hazard (De Meza and Webb 1987) and adverse selection problems (Ackerlof 1970).

It has been proven that equity and credit rationing occurs in markets with strong asymmetric

information (Stiglitz and Hellman 1998 and Myers and Majluf 1984). This also affects equity

capital. As we wrote in Section 1, the mix of external equity and debt sources will be affected by

differences in the relative severity of these information problems.

Jensen and Meckling (1976) raised the problem of agency costs. They said that the operating

management of firms involves a number of contractual relationships with the various stakeholders

in the company (owners, managers, customers, suppliers, and creditors). All these relationships

imply agency costs deriving from the potential conflict of interests that may arise between them,

and from the possibility of any of the stakeholders extracting private benefits from the firms,

compromising its economic efficiency in pursuit of the maximization of its value. For this reason,

the optimal financial structure is the one that minimizes the agency cost associated with debt in

relation to its benefits. Myers (2001) suggests that conflicts between managers and investors can be

solved by the latter taking on some responsibility for monitoring company performance, and by

incentive systems that link managers’ remuneration with firm performance.

An important contribution to the topic of the financial structure of firms was made by Myers

and Majluf’s (1984) pecking order theory (POT) and signalling theory. They suggest that because

of information asymmetries, only managers know the real value of the firm and its investment

opportunities. Investors do not know the real distribution of company income. This affects the

market value of the company, which can be over- or under-estimated, depending on the information

available in the market. Since managers will only decide to issue new stocks for investors if they

52/101

think that they are over-valuated by the market, this strategic decision will be considered a ‘bad

signal,’ which will cause a reduction on the market value of the firm. Dierkens (1991) highlights

that this shrinkage is more pronounced if there are more problems of information asymmetry

between managers and investors. In order to avoid this, managers will seek finance if their firms

have instruments that limit information asymmetries. If there are sufficient internal funds, every

project that should create value (i.e. has a positive forecasted NPV) will be undertaken. This is

unlikely in the opposite case, because, to avoid penalizing the market value of the firm, managers

could decide not to ask for capital in the market, even if this implies abandoning potentially

profitable projects. For the same reason, they may decide to subscribe debt rather then equity

capital, once internal funds are extinguished, as it is less influenced by information asymmetries.

Myers himself (1984) extends this theory, establishing that companies confront investment

opportunities and financial needs in a hierarchical order. Internal funds will be preferred, followed

by debt capital and, at a lower level, the issue of equity.

At first, POT and most of the other theories we have discussed were formulated to explain the

financial strategies of large quoted firms. These companies are characterized by a clear separation

between control and ownership, which is divided among a large number of shareholders, and by the

possibility of using stock or bond issuing as a source of financing. Several authors have argued that

POT is also valid with SMEs (Ang 1991, Holmes and Kent 1991, Cosh et al 2005 and Cosh and

Hughes 1994).

According to traditional POT, debt finance is preferable to equity; recent research, however, has

shown this pecking order can be reversed when investors’ skills and value-added are relevant

Garmaise (2000). Sau (2007) and Paul, Whittam and Wyper (2007) suggest that the traditional

hierarchy can be inverted with innovative firms when VC performs the evaluation function

(reducing the extent of information asymmetries) after which innovative firms can turn to bank

credit (revised POT).

Which of the two paradigms (POT and revised POT) explains the behaviour of NTBFs better is

strongly debated in the literature. Ǻstebro (2002) developed an empirical analysis of start-ups’

search for bank funding, finding that a large portion of them do not even apply even though a

significant proportion of applicants do receive loans. Schäfer et al (2004) use project and financial

dimensions to proxy the overall risk. They analyse the likelihood that NTBFs will obtain external

financing, looking at size, assets, project and novelty. According to their findings, risk does not

have a predictive power concerning the likelihood of a company receiving debt or equity. Carpenter

and Petersen (2002) say that, for small hi-tech firms, equity financing has some advantages

compared to debt: first, risk capital does not require any collateral to guarantee the funds conceded;

second, returns for investors are potentially superiorly unlimited, representing an incentive for VC

53/101

funds to invest money and time in potentially highly profitable firms; third, equity financing does

not increase the probability of company default and consequent bankruptcy costs.

Audretsch and Lehmanns’ (2003) findings show that the likelihood of obtaining venture capital

is inversely related to the extent to which the firm is financed by debt, as venture-backed firms have

significantly less debt. Therefore it appears, consistently with the greater part of the literature, that

equity provided by venture capital appears to be a substitute for debt rather than a complement to it.

Cosh et al (2005) produced a very interesting piece of research on the role of POT in the

financial strategy of SMEs. Although the sample was drawn from generic SMEs, not NTBFs, they

found the most significant determinant of applications for external finance to be the ratio

capex/profits, indicating support for the traditional pecking order whereby firms finance new

projects internally before seeking external finance. As we said earlier, the issue is no longer whether

or not firms follow a defined hierarchy of funding source; it is which of the two – POT or revised

POT – is adopted by NTBFs. Cosh et al found that start-ups with higher capital expenditures and/or

profits are more likely to obtain capital from a bank than from VC funding. In this paper we try to

solve this dilemma, borrowing Cosh et al’s approach (i.e., using capital expenditure and profit

profiles as the first determinants of the financial structure of a NTBF).

We also control for the role played by the size of the firm at the start up, a dimension often

investigated in the literature in relation to the funding strategy of NTBFs. As we noted earlier,

Schäfer et al’s (2004) empirical analysis failed to find a robust risk measure to predict access to

debt and equity for NTBFs. There is an interesting exception, however, concerning the indicators of

financial risk. The amount of money needed for a project is a powerful predictor of whether the

project will be equity financed. In Minola et al (2007), we showed that size of investment seemed to

have the highest predictive power of NTBF financing choice: larger amounts of money are likely to

be funded by venture capitalists or business angels, smaller amounts by a bank loan.

Similarly to Ueda’s (2004) predictions (larger projects are financed by VC), Sau (2007)

suggests that, although VC enhances the efficiency of the financial system, it is not suitable if

projects are at an embryonic stage and call for limited financial resources. Marked scale

diseconomies in management and monitoring discourage VC financing. We therefore expect a

stronger relation between size and VC financing than between size and debt financing. We proxy

the size of the firm with number of employees and turnover.

R&D intensity, technological uncertainty, innovativeness and asset intangibility constitute a

third, closely examined group of determinants of NTBF financing. According to Cosh et al (2005),

firms that have recently developed an innovation are approximately 35 per cent more likely to seek

and obtain capital from a VC fund. Higher asset intangibility is associated with risk (Garnsey 1995),

54/101

information asymmetries, agency costs and increased costs of external finance, both debt and

equity.

Collateral (all the assets a business can pledge as a guarantee for a loan) is often the principal

determinant for access to debt capital. The literature argued unambiguously that entrepreneurs with

little collateral tend to go to VC (Ueda 2004). We will check for the relevance of a well established

scheme of public guarantee to debt, like SFLG (described earlier).

The impact of market competition (Porter 1998) on a firm’s financing decisions has been

identified (Shepherd et al 2000). Profitability tends to be lower in more competitive industries

(especially when competitors are larger and dominant), which makes these companies less attractive

to investors. Firms that face more and bigger competitors are less likely to obtain the amount of

external capital they seek (Cosh et al 2005). We expect this phenomenon to be more pronounced for

obtaining equity, since it could be argued that banks are eager to finance a company in a highly

concentrated market, where the number of competitors proves the business idea is viable and the

project sustainable.

Finally, we control for the way the business originated. Cosh et al (2005) say that brand new

start-ups and businesses founded to implement an invention are more likely to be equity financed.

Business that originate within universities are referred to as university spin-offs (USO) and can

be viewed as a subset within the broader category of NTBF start-ups (Tang et al 2004). For the

purposes of our research, USOs are defined as new companies created by an academic/researcher

who forms a new company to exploit the outputs of university research and where the university

retains an investment in the company (typically in return for licensing the right to use university-

owned IP). USOs represent one specific way in which universities disseminate technologies and

deserve particular attention as some of the most relevant players in promoting innovation and

technology transfer.

USO management teams may lack a proven commercial track record (Steffensen et al 2000),

and university technology transfer offices are often short on critical business development and

business venturing skills (Vohora et al 2004). These USOs may have long gestation periods and

substantial resource requirements and may take many years to reach the market with products or

services that produce revenue. In addition, many USOs operate in new, emerging and sometimes as

yet undefined markets. The convergence of these factors and other variables leads to a significant

proportion of USOs being undercapitalized and deficient in the requisite knowledge and social

capital to access and acquire financial capital for development (Vohora et al 2004).

According to the taxonomy of USO developed by Heirman and Clarysse (2004), where parent

organization, business model and available (financial and technological) resources are matched with

possible financers, it is more likely that USOs will access equity capital.

55/101

Conversely we do not expect NTBFs established as a spin-offs from existing businesses or

M&A/MBO activity to be capital rationed, since the former are accompanied by collateral and

resources from mother companies and the latter by definition are accompanied by large funds and

the banks that support the entrepreneur

DATA

The sample we selected was retrieved from the UK Data Archive, database SN 4431,

Cambridge Centre for Business Research SME Dataset (second panel, 1997) – the same database

used by Cosh et al (2005). It is a longitudinal panel database of SME data relating to a wide range

of non-financial and attitudinal characteristics, and a limited number of financial variables not

normally available in modified company accounts, taken from a national postal survey.

The database forms the beginning of the second panel; the first panel was started in 1991. A

postal survey was sent to over 10,000 independent SMEs in the manufacturing and business

services sectors in England, Scotland and Wales. Just over half of the firms (5,430) were contacted

by telephone before the questionnaire was sent out to them, and 4,640 firms were sent the

questionnaire blind.

The questionnaire covered the following topics: general business characteristics; workforce and

training; commercial activity and competitive situation; innovation; factors affecting expansion and

efficiency; acquisition activity, capital expenditure and finance. A summary of the most relevant

information on the database is presented in Table 1.

Table 1 Information about the dataset (source: www.data-archive.ac.uk and Cosh et al 2005)

Country Great Britain Coverage (time period covered)

1994–1997

Dates of fieldwork June 1997–September 1997 Population Independent SMEs in Great Britain with fewer than 500 employees in

1997 Time dimensions Cross-sectional (one-time) study Sampling procedures One-stage stratified or systematic random sample Number of units 2,520 (obtained) Method of data collection

Postal survey

Date of release First edition 12 December 2001 Copyright Copyright University of Cambridge, Centre for Business Research

56/101

The original dataset was focused on SMEs in general, whereas our interest is explicitly focused

on NTBFs; therefore it was necessary to filter the original dataset, with regard to firm age and

industrial sector, in order to obtain a sample that would be relevant to our research.

NTBFs are normally less than 25 years old. Since our research explicitly focuses on the

financial issues linked to the initial phases of the firm lifecycle we concentrated on firms that were

less than six years old (the time taken, on average, to run the three initial phases: seed, early stage

and start-up). As to industrial sector (classified in the dataset through the SIC 1992 code), we

selected high-tech sectors – it is not an easy definition and there are differences in definition and

taxonomy in the literature. We adopted the definition of high-tech derived from the Greater

Cambridge Annual Profile 200711 and eventually obtained a sample of 303 NTBFs.

The sample is representative of NTBFs in the UK, since the original dataset was obtained

through a sample selection process. Detailed information on the methodology used to build the

sample can be found in Bullock and Hughes (1998).

METHODOLOGY

An original contribution of our research is our use of an expert system to shape the model of

NTBF fundraising processes (expert systems are used not to create theory but scientific

knowledge); in particular we used the methodology of Bayesian Network, also called Bayesian

Belief Networks (henceforth BBN), which is a compact model representation for reasoning under

uncertainty, based on the computation of the probability of different events or hypotheses given

number of observations.

G  P 

   P(X2|X1)    P(X3|X1)    P(X4|X2, X3)  

Figure 1 BBN: nodes, links and conditional probabilities

A Bayesian network N = (G,P) consists of a qualitative part (G), which is the direct acyclic

graph (DAG) structure G = (V,E), made by a set of nodes V and a set of directed edges E between

11 www.gcp.uk.net/downloads/G_C_Profile_07.pdf

57/101

nodes, and a quantitative part (P), which is the set of conditional probability distributions. Each

node has a finite set of states; attached to each node X4 with parents X1…Xn there is a conditional

probability table (CPT), which provides P(X|X1…Xn). Figure 1 shows a simple example of a BBN.

There are many advantages of expert systems; among others, they provide consistent answers

to repetitive decisions, processes and tasks, and represent significant support for decision making.

Within the broad field of expert systems, general rule-based systems have a series of problems. For

example, it requires substantial effort to construct and maintain them; they do not represent causal

knowledge explicitly; and numerical representations of uncertainty may be unsound. BBNs can be a

powerful tool, since they model the problem, not the expert, domain and support – rather than

substitute – the expert. BBNs use classical probability calculus and decision theory, not an

incoherent uncertainty calculus. This can be a great advantage, compared with fuzzy logic: the

robust statistics that back BBNs help to avoid the so called ‘black box’ risk, i.e. rules and

conclusions that have no strong evidence for their meaning). Furthermore, every type of variable

can be processed since BBNs allow the treatment of ordinal, binary and qualitative variables.

A BBN can represent the topology of a problem clearly and intuitively, within a robust

mathematical framework. Both the structure and the numerical parameters of a BBN can be elicited

from an expert. They can also be learned from data, as the structure of a Bayesian network is simply

a representation of dependences in the data; the numbers are a representation of the joint probability

distributions that can be inferred from the data. This represents a significant advantage, as it means

BBNs can help shape complex and holistic models, where multiple variables are mutually

interrelated and statistical dependent.

BBNs are probabilistic networks, capturing a set of (conditional) dependences and

independences associated with variables in the network; they get their name because the

computation of the probability from both data and knowledge is made on the basis of Bayes’

theorem.

There are three-steps to the process of building BBN learning from data: data discretization,

structure learning and parameter learning. Each step can be performed in several ways using

different parameters and algorithms, not necessary to detail here. Once our data had been cleaned

(or pre-processed), we proceeded with the linear discretization process, since BBNs work with

discrete values for the variable (there are continuous BBNs, but the complexity of the problem

required a more detailed exercise). Each continuous variable had to be segmented in groups of

intervals, whose number and width are set by the user, whereas the ordinal variables (such as Likert

scale answers), dummy variables and label variables are by definition discrete.

We performed structure learning with two pieces of software. The first was GeNIe, a

development environment for building graphical decision-theoretic models developed at the

58/101

Decision Systems Laboratory, University of Pittsburgh; the second was Hugin, probably the most

used management and industrial processes control software in the world. Both contain

comprehensive, flexible and user-friendly graphical interfaces, with an editor, compiler and runtime

system for the construction, maintenance and usage of knowledge bases using Bayesian network

technology.

We used Hugin to build the network, since it allows a very flexible process of structure

learning; in particular it provides p-value for each connection and allows the user to set a threshold

on the p-value, below which not significant links are disregarded. Hugin is also smart for parameter

learning as the user can add and drop links and the software recalculates CPTs very quickly.

Conversely GeNIe has a more friendly interface and performs the simulations process and

sensitivity analysis very well on existing networks.

As to structure learning, the user is asked to formalise prior knowledge on the topology of the

domain. The software then calculates the significance of all possible connections through the K2

algorithm, plotting a provisional framework with user-set arcs and some others that are suggested

because they are significantly supported by data. At this point the user can accept or modify the

network.

Once the network is built, parameter learning can be performed. The software calculates the

CPT that is subsequently provided for each node. After this the network is ready and the user can

play with it, setting some evidences on one or more nodes of the framework and observing how this

affects the joint probability of other nodes (this is known as propagation).

DATA ANALYSIS

In this section we describe an empirical analysis of a sample of 303 UK NTBFs, in which we

aimed to shed new light on the role of the determinants detailed in Section 2: capex, profit, reason

for firm creation, R&D activity, collateral (SFLG), size and market competition.

Variables definition is illustrated in Annex 1. In Table 2 we report some descriptive statistics

about the sample that provide some interesting insights. Fifty-one per cent of firms in our sample

did seek external finance, a high level that confirms pecking order theory: given their low ability to

generate internal finance (retained earnings), NTBFs are under more pressure than SMEs in general

to target external finance providers. Firms also managed to secure – on average – 80 per cent of the

external capital sought; this may not be fully explained by banks and VC provision, since the

dataset also investigates other forms of funding (leasing, factoring, invoice discounting, etc.)

The first significant surprise comes from the access to bank loans: 41 per cent (=1–0.59) of

firms approached banks and 80 per cent (=0.33/0.41) managed to secure some debt funding, which

59/101

is in sharp contrast to the stereotype that NTBFs do not (or cannot) access debt, although this may

be mitigated by a self selection bias. In particular, a large proportion of the samples (53 per cent,

=0.26/0.41) managed to secure the whole amount they were seeking from a bank.

There seems to be low credit rationing, since approximately only eight per cent of firms have

been rejected by banks (19 per cent of those who applied for a bank loan); the rejection rate from

VC is higher (three per cent, that is approximately 43 per cent of applicants), which is not

surprising.

Table 2 Descriptive statistics

Variable Name Not approached

Approached and rejected

Approached and financed

(some)

Approached and financed

(full) Observations

External finance sought from banks 0.590 0.080 0.070 0.260 290

External finance sought from VC funds

0.930 0.030 0.010 0.030 290

Variable Name Mean Median Standard Deviation Observations

External finance sought 0.510 1.000 0.501 298

Amount sought 416.826 80.000 1415.632 132 Percentage obtained 80.408696 100 35.013756 115 External finance obtained from banks 49.081 50.000 39.589 99

External finance obtained from VC funds

35.765 25.000 39.919 17

R&D last year 0.493 0.000 0.501 272 R&D expenditure 47.787 0.000 333.970 239 Turnover 2189.458 408.500 8432.829 262 Employees 25.778 7.000 46.847 284 Profit 130.425 35.000 386.281 241 Capital expenditure 441.505 25.000 3767.279 234 SFLG 0.137 0.000 0.345 270 Total competitors 29.610 4.000 311.589 267 Larger competitors 27.101 3.000 293.180 238 Overseas competitors 0.980 0.000 2.462 249

Variable Name Never Occasionally Continuously Observations R&D integration 0.350 0.350 0.300 276

Variable name MBO or M&A

Business spin-off

University spin-off New start-up Observations

Reason for firm establishment 0.250 0.250 0.060 0.440 294

60/101

Firms on average have a turnover close to £2 million, 25 employees, are profitable (£130,000)

and have capital expenditure of about £400,000, which is similar to the amount of funding sought.

These are considerable amounts for start-up companies and may be explained by the large

proportion of corporate and industrial spin-offs in the sample – larger-scale when they start – and by

the presence of a high number of manufacturing NTBFs, which are highly capital intensive.

The vast majority of firms perform some R&D, but only 30 per cent are engaged in continuous

R&D activity; 49 per cent did R&D in the previous year but the average R&D expenditure appears

pretty low (only 2.18 per cent of turnover).

SFLG schemes are accessed by 14 per cent of firms, which indicates this public scheme as a

successful initiative. Finally, market competition has few entrants, almost exclusively larger and

with a domestic provenance.

In Annex 2 we represent the correlation matrix, calculated on both dependent and independent

variables. It seems no multicollinearity phenomenon is present among the variables in the sample,

since no theoretically relevant correlation coefficient is found to be strong, but we can gather some

interesting evidence from the correlation analysis. In particular, it appears that continuous R&D

activity is correlated with smaller, more fragmented but more internationalized markets (see

correlation with Total competitors, Larger competitors and Overseas competitors), which is

unsurprising for new-technology based markets.

Judging by numbers of employees, all kinds of spin-offs and independent start-ups are smaller

in size, whereas firms that grow out of MBO and M&A activity are strongly correlated with larger

size. As concerns profit and capex, independent start-up results are negatively correlated with

profitability and capital intensity, whereas firms starting from MBO have strong and positive

correlation. Neither of these results is surprising.

Finally, accessing SLFG schemes is surprisingly positively correlated with performing

continuous R&D activity, which may indicate that the scheme targeted innovative firms, although

the theoretical prediction that university spin-offs are unlikely to target debt sources and access

guarantee schemes is confirmed.

The next step for the data analysis consisted of creating BBNs (described in Section 4). In this

paper we analyse the impact on access to debt and equity capital of a given subset of independent

variables: capital expenditure, profit, R&D activity, collateral (SFLG), size (turnover and

employees) and market competition. We preloaded a network where each independent variable was

linked to both dependent variables (see Annex 1). The software first returned the significance of

each link (dependency) through the p-value calculation (see Table 3), then suggested links among

independent variables that were not preloaded.

61/101

Most statistical dependencies were found to be strongly significant, except for impact of R&D

activity and capital expenditure on VC financing and – most surprisingly – the impact of the

number of competitors on both debt and equity financing. These non-significant links were deleted.

Table 3 The statistical significance of dependencies between input and output variables

p-value External finance

sought from VC External finance

sought from banks Reason for start-up 0,0742 0,0267

Employees 0,005 0,0001

Turnover 0,04 0,0075

Profit 0,0383 0,0435

SFLG 0,0089 0,0294

Capex 0,1481 0,015

R&D activity 0,6024 0,042

Competitors 0,5668 0,5019

Figure 2 BBNs derived from data

Our empirical findings point out conditional dependencies between profit and turnover, turnover

and capital expenditure, employees and turnover, and employees and reason for establishing the

firm. We want to stress the importance of this feature of BBN methodology. The discovery and

construction of the topology of the domain is an extremely important feature of BBN methodology.

62/101

It captures indirect dependencies that univariate or classic direct inference may not be able to show.

The final network is represented in Figure 2.

As we indicated earlier, once the network had been built, the software (specifically GeNIe)

allowed us to play with it, i.e. to set one or more evidences on some nodes and observe the impact

on other nodes and the network as a whole. We set three sample cases, illustrated in Table 4, with

relative impact on the two output variables (probability distribution of debt and equity financing).

Table 4 Summary of the examples of propagation

Variable Case1 Case2 Case3

Evid

ence

s set

on

inde

pend

ent

varia

bles

Reason for start-up New start-up Employees <5 <5 Turnover <100 <100

Profit SFLG No Yes Capex 100<x<500

R&D activity Continuous

Effe

cts

on d

epen

dent

var

iabl

es

B

- Not approached: 25%

- Rejected: 23% - Offered less: 23% - Offered full: 29%

- Not approached: 46%

- Rejected: 17% - Offered less: 20% - Offered full: 17%

- Not approached: 31%

- Rejected: 22% - Offered less: 22% - Offered full: 26%

VC

- Not approached: 37%

- Rejected: 21% - Offered less: 21% - Offered full: 21%

- Not approached: 71%

- Rejected: 10% - Offered less: 10% - Offered full: 10%

- Not approached: 71%

- Rejected: 10% - Offered less: 10% - Offered full: 10%

Case 1 was set with features typical of a company likely to seek VC funding, according to

theoretical observation: a high level of R&D activity, a completely new start-up with no parent

organization providing support and collateral, and a considerable amount of funding needed

because of the high level of capital expenditure. As a matter of fact, with respect to average

distribution of output node probability, we can observe an increased likelihood to access equity

financing, comparable to likelihood to access debt. Cases 2 and 3 were as small-scale start-ups; in

these cases, as expected, the likelihood of obtaining debt financing is much higher than the

likelihood to getting VC backing. Interestingly, debt is much more viable when a public guarantee

scheme is available as collateral (as in Case 3, compared to Case 2, where it is not available)

whereas the likelihood of receiving VC backing does not change at all.

Case 1 is also graphically illustrated in Figure 3; the nodes whose values were set by the user

can be identified through a unique symbol at the bottom right, whereas the ‘propagated’ nodes have

a solid box in the same position. Finally, we performed a sensitivity analysis in order to observe the

impact of variation in independent variables (input) on the two dependent variables (output) and,

because of the topology of the network, on the overall structure.

63/101

Figure 3 Examples of BBNs after propagation

Changing one input variable at a time, we analysed the variation in probability of the CPT of the

output nodes. To determine the relation we simply searched for monotone positive or negative

variations of probability distribution in output along with monotone positive variation in input. The

intensity of the relation was determined by the slope of the variation. In some cases it was not

possible to define univocal relation (see ‘/’), whereas non-significant relationships are defined as

‘ns’.

Results are summarized in Table 5, together with expected results on the basis of the theoretical

analysis developed in Section 2.

DISCUSSION OF THE RESULTS

We have only a partial confirmation of pecking order theory since Capex, as expected, raises the

likelihood of being external financed (both sources) but no univocal relation is found between

financing sources and Profit. We also tested for the ratio Profit/Capex (on a different BBN not

illustrated here), finding no significant relation with VC financing and a significantly positive

relationship between the ratio and access to debt. These results, together with observations based on

descriptive statistics, suggest that the extreme financial needs of NTBFs, especially with high

64/101

capex, force the appeal to external sources of finance. However, no hierarchy can be revealed on the

bases of profitability as a measure of a firm’s ability to generate internal funds; i.e. it does not

appear to be possible to distinguish as a separate segment which NTBFs adopt the POT rather than

revised POT paradigm on the basis of the simple operationalization we have proposed.

Table 5 Results of sensitivity analysis

B VC

Expected Sensitivity analysis Expected Sensitivity

analysis Profit ++ / + +

Capex + + + + Reason for start-up

MBO or M&A + + + +

Business spin-off + / + +

University spin-off - - + +

Independent start-up - / - -

R&D activity

Occasional - + + ns

Continuous - - + ns

Employees + + / ++ Turnover + + / ++

Competitors + ns - ns

Overseas competitors + ns - ns Larger competitors + ns - ns

SFLG + + / /

As far as reasons why the firm was established are concerned, as expected USOs are prevented

from receiving debt and are likely to be financed by a VC (Heirman and Clarysse 2004) because of

the high level of knowledge and expected growth associated with the low level of managerial

attitude that can be reduced by investors’ added value. Independent start-up is the only type for

which it is difficult to receive equity financing and represents the stereotype of businesses that have

to bootstrap to secure resources. Business that begin within a larger financial operation (spin-offs,

MBOs or M&As) are more likely to secure money from VC, which is coherent with our discussion

in the literature review above.

The level of R&D activity is not found to have a statistically significant relation to equity

financing. Continuous R&D activity, as expected, is negatively correlated with debt financing, but

65/101

this does not hold true for firms involved in only occasional R&D. This may once again indicate

that debt financing must be seen as a determinant player in fostering entrepreneurship: ‘Bankers

must be prepared to look as much to the future cash flow and maturity of the new technology

company as to security and gearing considerations […] But are technology firms really more risky

than other small companies? Or is it just that the banks believe that tech firms are riskier and so do

not proceed with appraising lending propositions in the tech sector?’ (Gill et al 2007)

Size is the strongest predictor of financing choice, with smaller NTBFs more likely to be debt

financed and larger ones likely to receive money from VC. This is expected and coherent with

theoretical predictions by Ueda (2004) and Schäfer et al (2004).

Finally, SFLG schemes have a positive impact on debt financing, whereas no significant

relation is found between NTBF gearing and the level of market competition (proxied by number of

competitors, overseas competitors and larger competitors).

CONCLUSION

‘Do firms, especially small, young firms, get enough money to fund viable projects? If not,

what, if anything, should be done about it? These questions are beguilingly simple to ask, and,

needless to say, much less easy for economist to answer’ (Cressy, 2002). The objective of our

research was to understand if debt is a viable source for NTBFs and to understand which factors

influence choice of sources of external financing.

We analysed factors at the firm level such as capex, profit, reason for firm creation, R&D

intensity, collateral (SFLG), size and market competition and found that, contrary to general

theoretical prediction, NTBFs do access debt. In particular, firms that start on a small scale, are

innovative but not extremely R&D focused, less capital intensive, and assisted by public guarantee

schemes have easy access to bank financing.

The likelihood of obtaining external financing increases with capital expenditure and is not

affected by profitability, R&D intensity and market competition, with the exception of bank debt,

which is higher for firms that perform occasional R&D activity and lower for firms that have

continuous R&D activity. NTBF funding is also influenced by the reasons why the firm was

established (university spin-offs and firms resulting from M&As are more likely to receive equity

financing than independent start ups) and by firm size (larger firms are more likely to get money

from venture capital funding). Many banks do take risks, but inevitably the latest financial crisis is

posing a severe challenge to the global banking system. If they are to continue fostering high tech

small businesses, banks will have to strengthen both their evaluation capability and accompanying

services in the start-up phase, which may benefit the growing entrepreneurial culture. Banks will

66/101

also have to provide personalised services and favour a shift in their business model, from a pure

service selling approach to one of more flexible risk-sharing with the entrepreneur.

From a theoretical point of view, our research results in weak acceptance of pecking order

theory; we suggest some empirical qualification of the pecking phenomenon by identifying so-

called accompanying (or mitigating) factors, i.e. variables at firm level that have clear and

significant roles in the fundraising process. We also found that, contrary to the larger part of

theoretical literature, banks can play a role in fostering technology-based entrepreneurship,

especially in an advanced context such as the UK, where the availability of public guarantee

schemes has proved effective.

From the methodological point of view, we believe our research enables an interesting holistic

analysis of the determinants of NTBF gearing, through the aggregation of quantitative, qualitative

and heterogeneous variables. Furthermore the use of BBNs, with significant dependencies and

topology elicited from data, help build a knowledge base that can be adopted to spread awareness to

policy makers, academics, practitioners and – most importantly – entrepreneurs on the determinants

of the fund raising process.

The use of expert systems is particularly useful when handling complex phenomena, where the

definition of variables and relationships is uncertain or unknown. Although numerous previous

studies, many cited in the literature review above, have focused on this topic and explored the

relationship of the variables with other methodologies, we think an expert system proved useful in

the context of our research. The variables examined were very heterogeneous and the network

collected them in a comprehensive topology. Furthermore many variables were qualitative,

subjective or nominal and therefore difficult to operationalize in a classic manner.

We acknowledge some limitations of the present work. First, we decided to design a model and

develop the sensitivity analysis on an example of BBN that takes into account only 10 independent

and two dependent variables. The BBN methodology allows for much larger models and more

complex topology to be handled, which we will consider in future studies. Second, we checked for

R&D intensity and public collateral schemes as indicators of knowledge intensity in firms, but did

not check for the role played by intangible assets that determine access to external finance; further

research could introduce dimensions such as patents, brand value and advertising intensity.

Interesting exploration may also be done on the role played by human capital, but this would

require moving from firm level analysis and approach to entrepreneur level. Third, we acknowledge

the existence of different development stages for NTBFs: seed, early stage and start-up firms may

represent very different organizations, with different financial dynamics and this can influence

results. We concentrated on generic fundraising actions occurring within the first three years of a

67/101

NTBF. Future studies might analyse the different dynamics underlying different stages within the

development cycle (Berger and Udell 1998).

Further research developments could include: more characterisation of debt contracts used by

NTBFs (what do they look like? collateral, impact on cash flow, repayment mode, etc.); the

extension of quantitative analysis on other country samples; a definition of optimality – in the

present work, accompanying factors have been identified on the basis of their occurrence in the

BBNs learned from data; no evaluation or inference can be made on the basis of any optimality

driver or performance measurement. Decision theory introduces a measure of preference, known as

utility (a function mapping the attributes of the possible outcomes of a decision process on a set of

real numbers), which can be implemented through a version of BBN, called influence diagrams.

This could help us derive from the knowledge base a predictive tool to guide entrepreneurs in start-

ups who are challenged by choices of financial strategy.

ACKNOWLEGDEMENTS

We are grateful to Francis Chittenden, Tim Minshall, Martin Holi, João Leitão, Annalisa Croce

and Khaled Soufani for useful discussions and helpful comments. We would also like to thank

seminar participants and discussants at the XXI RENT conference in Cardiff, UK, T2S Conferences

in Palm Desert, CA and Albany, NY, and at the CTM workshop in Cambridge, UK. We

acknowledge the support of Fondazione Politecnico di Milano.

REFERENCES Ackerlof, G.A. (1970) The Market for ‘Lemons’: Quality Uncertainty and the Market Mechanism, The

Quarterly Journal of Economics, 84(3), 488-500. Acs, Z.J. (2004) The value of entrepreneurial start-ups to an economy, Seminar Discussion Paper, Diebold

Institute for Public Policy Studies. Ang, J. (1991) Small business uniqueness and the theory of financial management, The Journal of Small

Business Finance, 1(1), 1-13. Astebro, T. (2002) Self-selection and bank loan screening on entrepreneurial ventures, University of

Waterloo, Working paper. Audretsch, D.B. (1995) Innovation and Industry Evolution. Cambridge, MA: MIT Press. In Audretsch, D.B. and Lehmann, E.E. (2003) Financing High-Tech Growth: The Role of Debt and Equity,

Working Paper, CEPR und Universität Konstanz, Indiana University. Berger, A., and Udell, G. (1998) The economics of small business finance: the roles of private equity and

debt markets in the financial growth cycle, Journal of Banking and Finance, 22, 613-673. Bullock, A. and Hughes, A. (1998) Survey design, response bias and sample characteristics in the 1997 CBR

SME survey (1998), in Cosh, A. and Hughes, A. (eds) Enterprise Britain: Growth Innovation and Public Policy in the Small and Medium sized Enterprise Sector 1994-1997.

Carpenter, R.E., Petersen, B.C. (2002) Capital market imperfections, high-tech investment and new equity, The Economic Journal, 112, F34-F72.

Cosh, A., Cumming, D. and Hughes, A. (2005) Outside Entrepreneurial Capital, Centre for Business Research University of Cambridge, Working Paper 301.

68/101

Cosh, A. and Hughes, A., (1994) Size, Financial Structure and Profitability: UK Companies in the 1980s, in Hughes, A. and Storey, D. J. (eds), Finance and the Small Firm, Routledge, 19-63.

Cressy, R. (2002) Funding gaps: a symposium, The Economic Journal, 112(477), F1-F16(16). De Meza, D. and Webb, D.C. (1987) Too much investment: a problem of asymmetric information, Quarterly

Journal of Economics, 102, 281-292. Dierkens, N. (1991) Information Asymmetry and Equity Issues, Journal of Financial and Quantitative

Analysis, 26(2), 181-99. Flash Eurobarometer (2005) SME access to finance, Flash Eurobarometer 174 – TNS Sofres/EOS Galilup

Europe. Fluck, Z., Holtz-Eakin, D. and Rosen, H.S. (1997) Where does the money come from? The financing of

small entrepreneurial enterprises, New York University working paper. Garmaise, M (2000) Informed investors and the financing of entrepreneurial projects, Working paper,

University of Chicago Graduate School of Business. Garnsey, E. (1995) High-technology Renewal and the UK Investment Problem, Journal of General

Management, 20(4), 1-22. Gibbons, M. (1994) The New Production of Knowledge, The Dynamics of Science and Research in

contemporary societies, SAGE. Gill, D., Minshall, T.H.W., Pickering, C. and Rigby, M. (2007) Funding Technology: Britain Forty Years

On. St John’s Innovation Centre and Institute for Manufacturing, Cambridge. Harris, R. (2001) The New Economy: Intellectual Origins and Theoretical Perspectives, International

Journal of Management Reviews, 3(1), 21-40. Heirman A. and Clarysse B., (2004) How and why do research-based start-ups differ at founding? A

resource-based configurational perspective, Journal of Technology Transfer, 29, 247-268. Holmes, S. and Kent, P. (1991) An empirical analysis of the financial structure of small and large Australian

manufacturing enterprises, The Journal of Small Business Finance, 1(2), 141-154. Holmstrom and Tirole (1997) Financial intermediation, loanable funds, and the real sector, Quarterly

Journal of Economics, 112, 663-691. Jensen, M.C. and Meckling, W.H. (1976) Theory of the firm: Managerial behaviour, agency costs and

ownership structure, Journal of Financial Economics, 3, 305-360. Minola, T. , Minshall, T. and Giorgino, M. (2007) Access To External Capital For Techno Start-ups, ISBE

Conference, Glasgow, November 2007. Minola, T. and Giorgino, M. (2008) Who’s Going to Provide the Funding for High-tech Start-Ups? A Model

for the Analysis of Determinants with a Fuzzy Approach, R&D Management, 38 (3), 335-351.

Modigliani, F. and Miller, M. (1958) The cost of capital, corporation finance and the theory of investment, The American Economic Review, 48, 261-297.

Myers S.C. and Majluf, N.C. (1984) Corporate financing and investment decisions when firms have information that investors do not have, Journal of Financial Economics, 13, 187-221

Myers, S.C. (1984) The Capital Structure Puzzle, The Journal of Finance, 39(3), 575-592. Myers, S.C. (2001) Capital Structure, The Journal of Economic Perspectives, 15(2), 81-102. Paul, S., Whittam, G. and Wyper, J. (2007) The pecking order hypothesis: does it apply to start-up firms?

Journal of Small Business and Enterprise Development. Pavlov, A., Poutziouris, P. and Soufani, K. (2004) Evaluating Flexibility in Small Business Finance, Journal

of Entrepreneurial Finance and Business Venturing, 9(1). Porter, M (1998) Competitive strategy: techniques for analyzing industries and competitors. Free Press. Rickne, A. and Jacobsson, S. (1999) New technology-based firms in Sweden: a study of their direct impact

on industrial renewal, Economics of Innovation and New Technology, 8, 197-223. Rothschild, M. and Stiglitz J. (1976) Equilibrium in competitive insurance markets: an essay on economics

of imperfect information, Quarterly Journal of Economics, 82, 629-649. Sau, L. (2007) New Pecking Order Financing for Innovative Firms: an Overview, Working Paper no.

02/2007, Department of Economics, Università di Torino. Schäfer, D., Werwatz, A. and Zimmermann, W. (2004) The Determinants of Debt and Private Equity Financing in Young Innovative SMEs: Evidence from Germany, Industry and Innovation, 11, 225-

248. Shepherd, D.A., Ettenson, R. and Crouch, A. (2000) New venture strategy and profitability: a venture

capitalist’s assessment, Journal of Business Venturing, 15, 449-467. Stam, E. and Garnsey, E. (2007), Entrepreneurship in the Knowledge Economy, Centre for Technology

Management Working Paper.

69/101

Steffensen, M., Rogers, E. and Speakman, K. (2000) Spin-offs from research centres at a research university, Journal of Business Venturing, 15(1), 93-111.

Stiglitz, J. and Hellmann, T. (1998) Credit and equity rationing in markets with adverse selection, European Economic Review, 44 (2000) 281-304.

Tang, H.K., Er Jackalin, Prasanthi Guda, Han, C.L. and Ong, H.W. (2004) Entrepreneurship in IC design in Singapore and Taiwan, International Journal of Entrepreneurship and Innovation Management, 4(3), 216-227.

Ueda, M. (2004) Banks versus Venture Capital: Project Evaluation, Screening, and Expropriation, The Journal of Finance, 59, 601-621.

Vohora, A., Wright, M. and Lockett, A. (2004) Critical junctures in the development of university high-tech spinout companies, Research Policy, 33(1), 147-175.

70/101

ANNEX 1: VARIABLES DEFINITION

Variable name Definition

External finance sought from banks

An ordered variable equal to 0 if a bank was approached but no finance offered, 1 if a bank was approached but offered less than the full amount, and 2 if a bank was approached and offered the full amount

External finance sought from VC funds

An ordered variable equal to 0 if a VC fund was approached but no finance offered, 1 if a VC was approached but offered less than the full amount, and 2 if a VC was approached and offered the full amount

External finance sought

A dummy variable equal to one if the firm attempted to obtain external finance (i.e. additional to internal cash flows)

Amount ought Thousand pounds

Percentage obtained The percentage of external finance obtained by the business in the 1996-1997 period (as a fraction of the amount sought)

External finance obtained from banks

The percentage of external finance obtained by the business in the 1996-1997 period (as a fraction of the amount sought)

External Finance Obtained from VC Funds

The percentage of external finance obtained by the business in the 1996-1997 period (as a fraction of the amount sought)

R&D last year A dummy variable equal to 1 if the firm did R&D last financial year

R&D expenditure Firm’s total R&D expenditures in 1996–1997. Measured in thousands of 1997 pounds

Turnover Firm’s total turnover in 1996–1997. Measured in thousands of 1997 pounds

Employees Firm’s total number of employees in 1996–1997

Profit Pre-tax profits (losses) before deduction of interest, tax, and directors’, partners’ or proprietors’ emoluments. Measured in thousands of 1997 pounds

Capital expenditure Firm’s total capital expenditures in 1996–1997. Measured in thousands of 1997 pounds

SFLG A dummy variable equal to one if the firm accessed the DTI SFLG support scheme

Total competitors Total number of serious competitors of the firm

Larger competitors The number of the firm’s primary competitors that are larger than the firm

Overseas competitors The number of the firm’s primary competitors that are overseas competitors

R&D integration Does the firm do R&D continuously, occasionally or never? Reason for firm establishment

How was the firm established? MBO, M&A, business spin-off, university spin-off or new start-up?

71/101

Annex 2 Correlation matrix

(i) (ii) (iii) (iv) (v) (vi) (vii) (viii) (ix) (x) (xi) (xii) (xiii) (xiv) (xv) (xvi) (xvii) External finance sought from banks

(i) 1

External finance sought from VC funds

(ii) 0,192172 1

R&D activity (occasional) (iii) 0,0246

8

-0,0415

1 1

R&D activity (continuous) (iv) 0,1661

27 0,1010

84

-0,4815

4 1

Dummy business spin-off

(v) -

0,02132

-0,0959

1

0,068382

0,003977 1

Dummy MBO (vi) 0,02671

0,169489

0,013875

0,01469

-0,1795

2 1

Dummy M&A (vii) 0,249303

-0,0536

7

0,022458

-0,0128

4

-0,1307

5

-0,1004

6 1

Dummy independent start-up

(viii) -

0,05968

0,027709

-0,1224

9

0,002402

-0,5136

1

-0,3946

2

-0,2874

2 1

Dummy university spin-off

(ix) -

0,11569

-0,0504

-0,0253

1

0,010047

-0,1227

8

-0,0943

3

-0,0687

1

-0,2698

9 1

Turnover (x) 0,1968 0,0073 - 0,1728 - 0,0823 0,0550 0,0262 - 1

72/101

45 43 0,07349

54 0,07866

51 13 59 0,05719

Employees (xi) 0,318914

0,132488

-0,0498

0,169572

-0,1044

2

0,41958

0,240143

-0,2227

2

-0,1123

4

0,636425 1

Profit (xii) 0,093777

-0,0051

5

0,033219

0,060819

-0,0642

4

0,285778

0,069254

-0,1284

4

-0,0747

9

-0,4357

2

0,213049 1

Capex (xiii) -

0,04296

-0,0001

9

-0,0772

1

-0,0491

3

-0,0614

4

0,293411

-0,0129

4

-0,1011

5

-0,0322

6

0,035892

0,150909

-0,0126

7 1

Total competitors (xiv) -

0,1615 -0,034 -

0,03227

-0,1187

8

-0,0385

7

-0,0329

8

-0,0017

8

0,052696

-0,0035

6

0,089769

0,042518

-0,0726

7

-0,0735

6 1

Larger competitors (xv)

-0,0277

4

-0,0748

6

0,10818

-0,2043

1

-0,0644

9

0,084792

-0,0724

0,019419

0,047112

0,012198

0,079254

0,046438

-0,0546

9

0,336164 1

Overseas competitors (xvi)

-0,0295

2

0,030451

0,029704

0,159875

-0,0639

4

0,233855

-0,0127

5

-0,0366

8

-0,0716

6

-0,0221

8

0,07905

0,142008

-0,0538

2

0,251968

0,175022 1

SFLG (xvii) 0,250747

0,288069

-0,0070

2

0,155424

0,008202

-0,0320

5

0,074159

0,076115

-0,1267

2

-0,0468

8

0,068459

-0,0050

4

-0,0507

5

0,036226

-0,1180

2

0,046985 1

73/101

EXTENDING THE KNOWLEDGE SPILLOVER THEORY OF ENTREPRENEURSHIP FOR ANALYSIS OF THE AUSTRALIAN

BIOTECHNOLOGY SECTOR

Julienne Senyard

School of Management, Queensland university of Technology, Gardens Point Campus, 1 George Street, Brisbane, Australia, 4000, Tel +61 73138 1010, Fax:

+44 7 3138 1313, E-mail: j.senyard@qut.edu.au

David Pickernell

Welsh Enterprise Institute, Business School, University of Glamorgan, Pontypridd, CF37 1DL, Tel +44 1443 483759 , E-mail: dgpicker@glam.ac.uk)

Nick Clifton

Centre for Advanced Studies in Social Sciences, City and Regional Planning, Cardiff University, Cardiff, Wales, CF10 3WA, Tel :+44 (0)29 208 76064, E-

mail:CliftonN1@Cardiff.ac.uk

Adrian Kay

Salford Business School, University of Salford, The Crescent, Manchester M5 4WT, Tel: +44 (0) 161 295 6902, Fax:+44 (0) 161 295 5022, E-Mail: Adrian Kay

A.Kay1@salford.ac.uk

Robyn Keast,

School of Management, Queensland university of Technology, Gardens Point Campus, 1 George Street, Brisbane, Australia, 4000, Tel +61 73138 1010, Fax:

+44 7 3138 1313, E-mail: rl.keast@qut.edu.au

ABSTRACT

Current research suggests that the process of knowledge creation is both networked and iterative. Synthesising the literature highlights a range of factors for analysis in knowledge-based industries. These factors are then used to examine the biotechnology sector in Queensland Australia, utilising available secondary literature, interviews with a range of broad stakeholders and 3 case–study companies. The results highlight issues regarding government policies for biotechnology, due to potential weaknesses in the network of relationships and governance between the key stakeholders (particularly within universities), the absence in some cases of relevant education (training and learning) for academics, and issues of entrepreneurial orientation and knowledge management in the use of created knowledge.

74/101

INTRODUCTION

The multi-faceted nature of innovation processes highlighted by Leyesdorff (2000)

suggests that this phenomenon involves a range of relationships (e.g. with other firms,

government agencies, universities) and learning and innovation occur through

interactive, iterative and networked approaches (Weick, 1990; Cooke, 1998). The

supply of such knowledge and its characteristics, also requires capable knowledge

users and effective knowledge transfer/translation (Cooke et al., 1997; Braczyk and

Heidenreich1998) to create commercialised outcomes in terms of product and process

innovation and improved firm capacity and growth.

The arguments surrounding this can be encapsulated within the knowledge spillover

theory of entrepreneurship. This argues, essentially, that knowledge developed in

some institutions might be commercialized by other institutions, and that

entrepreneurship is one way that the ‘economic agent with a given endowment of new

knowledge’ can best appropriate the returns from that knowledge (Acs et al. 2004).

Audretsch and Lehmann (2005) demonstrated, for example, that the number of new

firms located close to a university is positively influenced by it’s knowledge capacity.

The complexity of knowledge intensive entrepreneurship, however, often creates

barriers to exploitation. This may be result from (1) failure of private firms and

public institutions to generate new knowledge; (2) failure of that knowledge to be

disseminated efficiently; (3) failure of individuals to exploit new knowledge; (4) a

range of other factors that make entrepreneurship difficult.

Spatial proximity also often positively affects knowledge spillovers from firms and

research organisations, reinforcing the asymmetric economic geography of prosperity

and accomplishment (Cooke et al, 2005). There is also evidence, however, that

knowledge-creating collaborations as well as disseminating mechanism can be non-

local in nature. A recent study into the effects of social capital found that both faster

growing and more innovative small firms tend to make greater use of non-local

networks (Cooke et al. 2005). In addition, Acs et al (2007) highlight that new

knowledge can be imported into a region through the activities of foreign

multinationals.

75/101

This paper, therefore, examine these issues, using the knowledge-intensive

biotechnology sector in Australia as a case study, due to its nature as a knowledge-

intensive industry, with clear knowledge spillovers potential in domestic and

international markets, and strong government policies at both national and regional

levels. The paper is structured as follows. First the literature surrounding the creation,

dissemination and utilisation of knowledge, and the role of geographical proximity

highlights a range of inter-related factors for analysis. Second, the methodology

section evaluates the biotechnology industry against these factors, to establish its

relevance, and the methods for analysing the Australian biotechnology industry

specifically are outlined. Third, the results from the Australian biotechnology industry

are then analysed in terms of knowledge creation, dissemination, and utilisation.

Finally, conclusions are drawn concerning future policy for this his industry, and the

potential focus for future research is discussed.

LITERATURE: KNOWLEDGE CREATION, DISSEMINATION AND

UTILISATION MECHANISMS AND THE IMPORTANCE OF PROXIMITY

Knowledge Creation: Government, Universities and Industry

Etzkowitz and Leydesdorff (1997) developed the ‘Triple Helix’ framework, arguing

that innovation creation occurs at the intersections between government, university

and industry. Frenz et al (2005) discovered, however, that the level of UK firm-UK

university cooperation is very low, concluding more generally, that firms must have a

certain level of absorptive capacity (defined by the proportion of science and

engineering graduates in the workforce, level of firm R&D expenditure, and

organizational capability) before entering into cooperation with a university. Once

established, however, this cooperation was found to have a positive and significant

effect on innovation. They also argued that the most consistent finding to come out of

regional total factor productivity growth studies was that the stock of human capital

enhances the absorptive capacity of firms, facilitating local technology transfer, local

and regional knowledge spillovers and growth.

In the knowledge-spillover theory of entrepreneurship (Acs et al. 2004), however, it is

also argued that levels of knowledge-based entrepreneurship might be affected by (1)

the ability of private firms and public institutions to generate new knowledge; but also

76/101

by (2) the degree to which this new knowledge is disseminated to the wider economy

as well as (3) the degree to which individuals and firms are able to exploit this new

knowledge. The absence of a domestic industry base and/or the absence of domestic

knowledge-creating institutions, such as public research institutes, might mitigate

against the emergence of knowledge-based entrepreneurship (Audretsch and Lehmann

2005), as might the absence of foreign multinationals in a region, able to import such

knowledge from outside.

Additionally, however, individuals or organizations with market knowledge or other

resources may not be aware of the new knowledge because of a lack of dissemination,

and therefore fail to invest, or under-invest, in the knowledge or in new firms

(Audretsch, 2004). Individuals may also fail to commercialise new knowledge via

entrepreneurship, if they underinvest in commercialization activities or fail in their

attempts to commercialize due to a lack of market knowledge, ability to manage the

new knowledge effectively or insufficient entrepreneurial ability. As regional

knowledge and innovation systems are dynamic and evolving, these issues can also be

affected by the nature of the region itself.

Knowledge Dissemination: Structures, Learning and Governance

The role of knowledge and its characteristics therefore also needs to be evaluated

through the lens both of the capabilities of knowledge users and effectiveness of

knowledge transfer/translation (Cooke et al., 1997; Braczyk and Heidenreich, 1998).

Links between entrepreneurial growth, innovation, and networking, for example, has

also led to an increasing focus on entrepreneurial firms networked together in various

ways (Asheim and Coenen, 2006), Gordon and McCann (2000) identifying three sets

of advantages in geographically based clusters and networks, derived from

agglomeration (i.e. from external economies of scale, scope and complexity).

Crucial here, are issues surrounding the management of networks, the structures and

forums in which the actors operate and the role of education and training and

processes of learning, to enable networked knowledge processes (Pickernell et al,

2008). There are a range of forums and structures, for example, in which and through

which knowledge creation, but also dissemination can occur, including direct spinouts

of companies, and collaborations with various stakeholder groupings from industry

supply chains, government institutions and universities. Cluster and network-based

77/101

approaches also suggest a range of other formal and informal mechanisms in which

knowledge creation and dissemination can also be encouraged. The suitability of the

structures and for a used, however, will be factors of crucial importance in

determining the success or otherwise of the knowledge creation and dissemination

process.

Encouraging the take-up of new innovations through dissemination via education and

training-based processes also allows individuals to be provided with knowledge about

the innovation itself, as well as being inspired and convinced of the possibilities for

success and mutual gain (Goffin, and Mitchell 2005). Such explicit, codified

knowledge can, however, be encapsulated in formats and transferred to users who are

able to interpret and utilise it independently from the context in which it was created,

(Howells, 2002). The transfer of codified knowledge is not seen as strongly dependent

on geography as codified knowledge can be transferred across geographic regions

fairly readily, and reductions in costs and improved communications increase access

to codified knowledge, rendering it less important as a source of competitive

advantage. Tacit knowledge, it has been argued, however, does not always travel

well, making it a key source of ‘the geography of innovation’ (Asheim and Gertler,

2005). This includes knowledge flows between firms, research organisations,

institutions and public agencies that are embedded in a regional context. Frenz and

Ougthon (2006), therefore, argue that, since proximity facilitates the transfer of tacit

knowledge transfer and learning - both of which are important determinants of

innovation - innovation activity takes on a strong regional dimension that may be

reinforced by agglomeration economies in production and pools of skilled

labour/human capital.

Effective and appropriate management of innovation creation and diffusion structures

and forums are also vital to this process. In this respect, the three basic modes or

mechanisms that can be applied are hierarchical state or corporation based, or the

market, or social networks (Lowndes and Skelcker, 1998). Markets are sometimes,

however, perceived as unable to adequately bundle the relevant resources and

capacities between science and industry, and the complete vertical integration

inherent in hierarchy restricts flexibility and incentives (Menard, 2002). Conversely

pure networks of relationships based on trust and reciprocity are often insufficient

forces to secure necessary directed outcomes (Rhodes 1997; Keast, and Brown 2002).

Hybrid approaches therefore have the ability to limit or balance out the negative

78/101

effects of an over-reliance on one governance mode (Menard 2002), through

exhibiting a number of possible combinations and re-combinations of contract and

trust to form effective strategic partnerships (Schaeffer and Loveridge 2002). Again,

the suitability of the management mechanism used in the situation at hand will be

important in determining the success or failure of the creation and dissemination

process, particularly given that there are a range of processes and motivations of

importance when examining these issues, depending on the nature of the network

being utilised.

The relationship between learning, structures and governance modes provides the

mechanisms to bring participants (and the various stakeholders) together to share

resources and knowledge that are present in individuals or organisations. One

scenario, for example, might see a myriad of key stakeholders from industry,

government, and institutions (including universities and government research

departments), utilising these interconnected mechanisms to generate and disseminate

knowledge, innovation, skills, and training, and to operate management and

governance structures appropriate to their own particular circumstances.

Knowledge Utilisation: Knowledge Asymmetry, Knowledge Management and

Entrepreneurial Orientation

In order to exploit knowledge created and disseminated through networks, there is

also the crucial role that entrepreneurship itself plays in the process. In particular,

entrepreneurs require appropriate personal ‘knowledge’, resources and management

abilities, encapsulated in the factors of entrepreneurial orientation, appropriate

strategic evaluations seen in asymmetric knowledge provision between existing

activities and the innovation, and appropriate knowledge management (Senyard,

2007). The issue of knowledge asymmetry, however, is a complex one as new

knowledge by its very nature creates knowledge asymmetry (or lack of proximity

between old and new knowledge – Boschma, 2005). This asymmetry cannot be too

large for firms receiving knowledge, however, as they will be unable to use the

knowledge received, and some symmetry of information is thus critical for

relationships development and success as it develops trust (Fukuyama, 1995,

Baranson, 1990), which positively affects decisions to maintain the relationship and

creates stability through shared understandings and norms. Knowledge asymmetries

79/101

exist because of differences in knowledge, business processes and resources

(Brooksbank et al. 2007). Cimon (2004) further evaluated and categorised

asymmetries as (1) information asymmetries; (2) knowledge asymmetries; and (3)

learning asymmetries, with all three recognised as having a role to play in the process

of organizational knowledge creation (Nonaka and Takeuchi, 1995; Ancori et al.,

2000), and arising from differing resource endowments (e.g. Barney, 1991) and

absorptive capacity (Cohen and Levinthal, 1990).

The way in which this new knowledge is then managed will also be crucial for the

strategic direction of the firm in many industries (Dyer et al., 2001). It is argued that

firms should be able to increase their competitive performance through effective

knowledge management, strategic learning and knowledge orientation, and that these

are positively related to long term survival and growth (Salojarvi et al 2005; Matlay

2000). Salojarvi et al (2005) also state that firm success often depends upon an

organisation’s ability to create, utilise and develop knowledge-based assets. Despite

this it is somewhat surprising that relatively few studies have examined the links

between knowledge management and firm growth, instead concentrating on

knowledge management alone (e.g. Kautz and Thaysen, 2001; Wickert and Herschel,

2001). Successful innovation requires managers to match ‘technical’ expertise, in

areas such as technology and project management, with ‘soft’ skills in people

management, to promote creativity. Few managers are either educated or experienced

in both of these areas (Goffin, and Mitchell 2005, p. 27). Kirby (2004) thus advocates

the use of Higher Education Institutions (HEIs) in promoting and reinforcing the

development of such entrepreneurial skills in communication, creativity, critical

thinking and assessment, leadership, negotiation, problem-solving, social networking

skills, and time-management. These are all equally applicable to the creation and

utilization of knowledge, the links with HEI also highlighting ways in which links can

be made between external network and internal-firm-based processes.

Effective entrepreneurial behaviour is also necessary to prosper in competitive

environments (Covin and Slevin, 1988, Lumpkin and Dess, 1996, Miller, 1983 and

Zahra, 1993a). Within this context entrepreneurship, plays a pivotal role in facilitating

links between research and industry (Abramson et al., 1997). Utilising Covin and

Slevin’s (1989) “basic uni-dimensional strategic orientation” concept a firm's

behaviour can be categorised along a continuum that ranges from highly conservative

to highly entrepreneurial behaviour in which a firm’s position is referred to as its

80/101

entrepreneurial orientation (EO) (see Barringer and Bluedorn, 1999; Lumpkin and

Dess, 1996). The three main dimensions of EO are innovation, pro-activeness and

risk. Previous studies (see Table 1) have consistently highlighted a positive

relationship between EO and performance.

Table 1: Dimensions of Entrepreneurial Orientation

The Issue of Proximity: local and Cross Local Linkages

Despite the importance placed on geographical proximity in the literature (highlighted

earlier for example in relation to tacit and codified knowledge), there is also evidence

that knowledge creation, dissemination and utilisation processes can also have a wider

geographical dimension. Boschma (2005), in particular, identifies five dimensions of

proximity that can have an impact on learning and knowledge, and which, crucially,

do not necessarily require geographical proximity. He argues that the need for

Author Dimension Sample Statistical Analysis

Study Conclusion

Lee, et al (2001)

Innovation, Risk Taking Pro-activeness

137 Korean Tech Start Ups

Regression, Correlation

EO has positive, marginally statistically significant effect on performance.

Voss et al (2005)

Innovation, Risk Taking, Pro-activeness, Competition Scanning, Autonomy

324 US Theatre Groups

Regression, Correlation

Relationship between stakeholder influence and EO behaviours is transparent, managers develop reciprocal, strategic relationships that reinforce valued behaviours. When the interaction between stakeholder influence and EO behaviours is less transparent, managers must perform a balancing act to contend with complex, pluralistic and conflicting stakeholder demands and responses.

Zhou et al (2005)

Opportunity Recognition, Environmental Sensitivity, Environmental change and challenges

China 350 respondents to brand

Structural Equation Modelling Factor Analysis

EO has a more positive impact on tech based and market-based innovation when competition is intense

Wiklund and Shepard (2005)

Pro-activeness, Innovativeness, and Risk taking

465 Swedish Manufacturing and services firm

regression analysis and correlation

EO positively influences small business performance. High EO, high access to capital, and environmental dynamism did not increase performance.

Poon et al (2006)

Innovation, Risk Taking, Pro-activeness

96 small firms Regression, Correlation

EO did not mediate the relationship between internal locus of control and firm performance. EO is a necessary mediator of the link between generalized self-efficacy and firm performance.

81/101

geographical proximity for learning to occur is weak when there is a clear division of

precise tasks that are coordinated by a strong central authority—organizational

proximity—and the partners share the same cognitive experience—cognitive

proximity (Boschma 2005: 69). He further suggests that spatial lock-in may be solved

or even avoided by establishing non-local linkages. Findings from several empirical

studies also suggest that non-local as well as local relationships are important sources

for interactive learning (see Asheim and Coenen 2006, Jaffe et al. 1993, Feldman

1994).

Boschma (2005) suggests that shared formal institution structures [such as laws, rules

and regulations that are the subject of governance] are not necessarily bound by

geographic proximity. Instead, institutional structures can reflect a kind of balance

between institutional stability (reducing uncertainty and opportunism) openness

(providing opportunities for newcomers) and flexibility (experimenting with new

institutions). To satisfy the need for co-presence to exchange tacit knowledge, cross-

location networks could also bring people together through, for example, occasional

travel (Boschma 2005). Asheim and Coenen (2006) argue, therefore, that there is a

need for both local and distant networks for effective process and product innovation.

This highlights the need, for factors related to knowledge (and its influence upon

innovation and growth) to be examined within firms’ entrepreneurial processes at

both local and cross-local levels.

Summarising the Factors for Analysis

Synthesising these multi-faceted relationships between knowledge, how it is

disseminated through the network, innovation processes and growth, creates a

comprehensive range of factors for analysis, which can be listed under the following

headings:

• Knowledge-creation relationships between firms, government and its agencies,

and institutions, such as universities).

• How knowledge-dissemination occurs though the structures for disseminating

knowledge (e.g. via spinouts, alliances, collaborative networks etc.),

management and governance of the relationships between the sets of

82/101

stakeholders, and the education, training and learning required for effective

dissemination.

• knowledge-utilisation by the companies themselves for innovation-specific and

more general growth-related outcomes, by examining knowledge asymmetry

(degree to which the knowledge is new / overlapping with existing

knowledge), its management and synergy with strategic behaviours and plans

that relates to entrepreneurial orientation in maximising the benefits of the

knowledge.

• Proximity issues by evaluating the role and importance of local versus cross-

local creation, dissemination and utilisation mechanisms

METHODOLOGY

In order to examine the absolute and relative importance of the identified factors,

however, there is a need for in-depth study within knowledge-based industries. This

requires a methodology that determines first a suitable industry for analysis, then a

suitable region / nation, to provide a case study setting. Lastly, relevant methods of

gathering the data are required.

Case-Study Industry Identification

Available literature recommends that owing to the myriad of stakeholders

(governments, between university and within university relationships, external

businesses including domestic and international partners, venture capitalists,

contractors etc), and the complexity in both knowledge seeking and knowledge

requirements, the biotechnology industry may be suitable for the in-depth study.

Governments assist in the development of a biotechnology industry through a policy

framework aiming to compensate for market failures (Orsenigo, 1989). Access to

resources and incentives, in particular financial resources from government, is of

great importance in any theory explaining biotechnology (Harman & Harman, 2004),

83/101

as without this, the level of R & D spending may be less than the optimum for the

economy (Erskinomics Consulting, 2003).

Knowledge in the biotechnology firm also often evolves as a result of synthesis of

scientific, technological and business knowledge, and managerial skills, these

knowledge systems coevolving as the firm develops (Liyanage & Barnard, 2003),

often occurring between university and within university relationships. Therefore a

firm’s capability is a continuous synthesis of scientific, technological and managerial

skills and knowledge requiring input from organisational learning and management

strategies (Lane & Lubatkin, 1998; Zahra & George, 2002; Zollo & Winter, 2002)

The advancement of the biotechnology industry is thus critically bound by knowledge

and information asymmetries associated with scientific, technological and business

related knowledge (Murray, 2002). Finally, in terms of the geography of relationships,

Cooke and Laurentis (2006) found, for example, that UK universities and firms in the

UK biotech sector, commonly enter into collaboration with overseas partners in

addition to partners within the UK, for product innovation, distribution, licensing

deals and supply contracts; and as such the UK biotechnology sector operates in a

global marketplace. With specific regard to innovation, as distinct from other

interactions such as research, joint patenting, purchasing or supplying, and other more

informal collaboration, the act of commercialising new knowledge in the form of a

product or service new to the firm or new to the market, Cooke and Laurentis (2006)

found that UK biotechnology firms also innovate collaboratively. In the absence of

distant spillovers from other sources, these firms often form collaborator relations

with ‘distant networks’ to augment R&D knowledge for themselves. These occur

broadly equally in the EU and North America, as well as more extensively in the

home country itself.

This suggests that the biotechnology industry generally exhibits the knowledge

generation, dissemination, utilisation mechanisms highlighted in the literature. In

order to examine these processes in more depth, however, the cases were developed

using Australian firms. Australia is often perceived to suffer from “smallness of

domestic markets” (Felsenstein and Portnov 2006) and owing to this, their spatial

proximity relationship profiles are of interest. This research and analysis occurred

within the 2005-2006 period, when Australia’s biotechnology industry was

experiencing growth. Market capital as at December 2005 for biotechnology, medical

devices and other healthcare companies whose stocks are listed on the Australian

84/101

Stock Exchange (ASX) is A$42.4 billion (157 companies) up from A$27.1 billion in

2004 (Australia Government Initiative, 2006). Partly as a result, Australia now has

more biotechnology companies relative to GDP than any country except Canada

(IMB 2005). In 2006, there were 427 core biotechnology firms with 625 firms in

medical devices (Ausbiotech 2007). The majority of these being small to medium

enterprises, recorded in the Australian biotechnology industry (Hopper & Thorburn

2005). There is, however, a relative paucity of research of this important industry in

the Australian context (Senyard, 2007), and in particular, an evaluation of the factors

of importance to government policy when seeking to encourage the industry’s

successful development and growth.

Table 2: Summary of Case Study Firms’ Background Data Title Case One (TT) Case Two (PB) Case Three (GT) SUMMARY DATA:

Type of Firm Start Up (R & D) Private Ceased Operating Fora for Knowledge

University Industry Incubator

Knowledge (IP) Patents Secured No Patent until later 8 Patents Product Definition

Platform is a patented, defined, set of protein complexes

Diagnostic technical platforms includes enzyme-linked immunosorbent assay (ELISA), indirect fluorescent antibody test (IFA) and rapid lateral flow devices

Blood-based monitoring tests for performance animals, initially the performance horse

Stakeholders Academic Scientist Director Commercialisation Manager

Scientist/CEO, Founding Members, Board of Directors

Academic Scientist, CEO, Board of Directors

University Role Positive Start Up, Continuing Relationship

Negative Prior Start Up Experience with University

Minor prior commercialisation role at university , Indirect use of university resources (knowledge) translated to own IP

Government Role (Funding)

Yes Yes (later) Yes

Board of Directors

Scientist did not want to be on Board

Scientist is currently on Board

Scientist wanted to be on the Board but did not

Strategic Orientation

Technology Driven: Target: Big Pharmaceutical (further funding)

Customer Driven, Target End customer (Hospitals

Technology/Market Driven. Target: Equine Industry

Data Selection Methods

85/101

A 3-stage data selection process was chosen to examine the issues in the Australian

context, with questions concerning proximity also built into each of these stages. First

analysis was conducted into available secondary literature sources, to establish the

knowledge creation environment. This was combined with (15) in-depth face-to-face

semi-structured interviews, to explore knowledge dissemination. The interviewees

were purposively selected (e.g. see Patton, 1990), according to their differing roles in

the management and governance, fora and structures, and learning processes which

surround knowledge dissemination. These stakeholders in the ‘triple helix’ came from

Industry (industry association Ausbiotech members, commercialisation managers,

venture capitalists, biotechnology employment specialists, entrepreneurial managers

of established start ups) Government (both Federal and State-level), and University

(science academics, commercialisation managers).

Finally, 3 case study companies (again purposively chosen, this time for their

divergent experiences, summarised in table 2) were also analysed in-depth, to

examine the role and importance of the internal firm-level activities in knowledge

utilization (commercialisation) processes.

The three case studies of firms show varying stages of firm development.

The first case study is still conducting clinical trials and is publicly listed on the

Australian Stock Exchange. It retains close links with its affiliated university (who

retains share ownership) and is, at present, looking for large global pharmaceutical

partners to continue product development.

The second firm, the most established of the three, has been very successful in the

global markets and continues to develop new products.

The third case study evaluates a business which no longer operates. The cases were

purposively selected to highlight different outcomes, product definition, age, and

provide a snapshot of knowledge utilisation behaviours (entrepreneurial orientation),

knowledge asymmetry, and knowledge management characteristics of the companies.

The three key individuals chosen for interview from each case-study company were:

the scientist who developed the science/technology and sought to commercialise the

product/service; the individual who provided the key relationship for business

information and processes during start up; and the current CEO.

The case-research again used a semi-structured interview guide on topics of

knowledge asymmetry, entrepreneurial orientation, relationship characteristics

including proximity themes and impacts on relationships developed.

86/101

Data Gathering and Analysis

Bothe the secondary literature review/industry interviews and cases including

interviews were used to examine knowledge creation, utilization and dissemination

processes. First, owing to confidentiality arrangements, the individuals and companies

are not explicitly identified in the results. The interview-based research in industry

interviews and case interviews developed protocol that included open-ended

questions to allow a natural conversation flow around the common set of issues

(Patton, 1987). The interviews conducted lasted between one and a half and two

hours. Individuals sought were first telephoned, explaining the request and research

being conducted. This was followed up by an e-mail document and request for

interview. A follow up telephone call the week after the e-mail was sent guaranteed a

high response to requests. Interviews were then conducted during the two weeks

following the follow-up telephone call where possible, in environments that were

suitable for the respondent, in most cases, their offices.

The interviews were recorded on mp3 audio. This was transcribed into a hard copy

format. Where necessary, the initial interview was followed up by a further interview

to clarify responses. A database of responses was then created, containing the

answers of all the interviewees to the specific topics. Analysis of the interview data

was then undertaken using the five stage analysis process suggested by McCracken

(1988). The first stage involves treating the statements in the interview on their own,

without making connections to other parts of the interview. In the next stage

observations are developed, individually, then according to the evidence contained

within the interview and then according to the literature. The third stage involved

interconnecting these developed observations by use of previous literature. The

interview schedule/transcript was then used only to check the ideas as the observation

comparison process is undertaken. The fourth stage involved collective analysis of the

collected, developed observations and statements, to subject them to analysis, to

investigate any consistent themes or inconsistencies which may be developed. In the

fifth and last stage, these themes and patterns were investigated through the interview

programme as a whole, to examine the themes amongst the interviewees.

87/101

Table 3: Official Knowledge Creation Strategies Package Overview Total

Funding and Duration

Key Knowledge Support Strategies

Backing Australia’s Ability (BAA I and II)

General science and Innovation package, focused on three key elements in the innovation process: • strengthening

Australia’s ability to generate ideas and undertake research;

• accelerating the commercial application of ideas; and

• developing and retaining Australian skills.

Total duration: 2001-11. Total Funding: $8.3billion.

Developing and Retaining Skills The package supports the long-term sustainability of Australia’s skill base in the enabling sciences and the encouragement of positive attitudes toward science and innovation in the community. It promotes this by: • Funding an extra 5740 higher education

places in ICT, mathematics and science at Australian universities ($350.5m)

• Improve teaching in Innovation, Science, Technology and Mathematics ($38.8m)

• Enhance capabilities of government schools to build stronger scientific, mathematical and technological skills of Australian students and to encourage school-based innovation ($373m).

• Questacon Smart Moves: an initiative to raise awareness of science and innovation among young Australians and encourage participation in science and innovation industries ($15.1m)

• Science Connections Programme: initiative to raise awareness of the contributions of science and innovation in the broader Australian community ($25.8)

National Biotechnology Strategy (NBS)

Provides a framework for the development of biotechnology in Australia. The strategy addresses six key themes: • Biotechnology

in the community;

• Ensuring effective regulation;

• Biotechnology in the Economy;

• Australian biotechnology in the global market;

• Resources for biotechnology;and

• Maintaining momentum and coordination

Total duration: 2000-08. Received initial funding of $30.5m in 2000, followed by additional contributions of $66.5m and $20m through BAA I and II.

HR for Biotechnology Development The key objectives are: • enhance management skills in the

biotechnology sector; • attract high quality researchers and

experienced leaders; • encourage entrepreneurship; and • monitor demand and supply for specialist

skills. The key strategies are: • Improve management of research,

intellectual property and technology within established firms and new enterprises;

• Develop, attract, motivate and retain high quality researchers, particularly in those fields where Australia has strong capacities to commercialize research outcomes;

• Maximize technological awareness and capabilities throughout industries that will be developing and applying biotechnology

• Develop programs and systems to foster entrepreneurship

• Monitor emerging skills needs in the biotechnology sector and develop appropriate responses.

Source: Stephens et al (2006)

88/101

RESULTS

Knowledge Creation: Overall Context

The results of the industry interviews conducted are now briefly outlined in terms of

examining the context and role of the 3 main sets of stakeholders (the industry and its

firms, government and university). In terms of the Australian biotechnology

industry’s stakeholders and knowledge creation policies, there has been a particularly

strong Australian federal government policy in place, in conjunction with university

institutions and the government, as highlighted in table 3.According to Stephens et al

(2006), Australian biotechnology, has a research strength underpinned by its

universities, its federal research body (Commonwealth Scientific and Industrial

Research Organisation (CSIRO), and other leading institutions. The majority of

Australian biotechnology firms in 2004 (60 per cent) are less than six years old

(Department of Industry Tourism and Resources, 2004) and the industry is developing

through small, dedicated entrepreneurial firms staffed mostly by scientists (Curtis et

al, 2006). Australian State Government initiatives have also shown an increased

focus on biotechnology research creation, dissemination and utilisation agendas.

Specifically, current programs developed by the Smart State Strategy (Queensland

Biotechnology Strategic Plan 2005) include:

• Smart State Innovation Projects Fund: Consists of $60 million over the next

four years to support national and international alliances and collaborations

between research organisations and industry.

• Biotechnology Commercialisation Pipeline: Assists new biotechnology firms

to access private sector finance and enables them to progress along the

commercialisation pathway.

• BioStart Fund: Provides access to early stage financing for startup firms.

• Current commercialisation training opportunities will be extended through the

• Mentoring for Growth and Innovation Start-Up Scheme program.

• Queensland Biocapital Fund: Through the QIC to l stage later venture capital

financing to ensure the establishment of globally competitive bio-businesses.

89/101

• The Government will raise investor readiness by encouraging participation in

the Commercialisation Bootcamp and Masterclass Program through the

Australian Institute for Commercialisation.

• Smart State Innovation Skills Fund: Providing A$ 12 million to attract and

retain leading scientists and build skills in Queensland.

The biotechnology industry provides a context of strong government support and

policy initiatives related to the creation and use of knowledge, with a strong potential

role for universities in developing the industry, but also with key governance issues

concerning how these knowledge processes can be managed and developed for the

most effective outcomes, given the different stakeholders involved, both domestically

and internationally.

Knowledge Dissemination Processes

Interviews with (15) key industry stakeholders indicated question marks over the

(strong) role of government hierarchical-based management in focusing university

agendas in particular, in biotechnology and commercialisation. As previously shown,

governments through financial provision to biotechnology programs also influence

knowledge dissemination processes through funding reporting requirements that

impact what type of information is disseminated, to whom and frequency of this

dissemination. An inability to satisfy these reports places doubt on future funding and

further development. Traditionally responsible for funding basic vs applied research,

governments are now shifting greater commercial responsibility on universities and a

focus on managing outcomes:

‘The main change has been really almost foisted upon universities by

government policy and that is the sense that they have to manage the outcomes

of their research, which was never ever something that was really on the

agenda in universities. Ten years ago it was that there would be papers

published, there would be a contribution to the academic arena, but there

would not be necessarily be any transfer of that information into commercial

value or into industry.’ JC

90/101

Unsurprisingly, therefore, management and governance conflicts occur between the

commercialisation agenda and the university bureaucracy and traditional social

agendas:

‘There are additional pressures of managing the conflicting roles within

academia. It’s actually very difficult to be entrepreneurial in a professional

bureaucracy and universities are the classic professional

bureaucracy…Essentially universities as professional bureaucracies have

really got to – well the mantra is that everybody is equal, okay. But in

business, everybody is not equal and that’s the dilemma that you have when

you’re an entrepreneurial scientist in a university.’ JC

The more-market based current approach to exploitation of university knowledge, in a

university model otherwise characterised by more hierarchy-based mechanisms also

highlights the key role of the commercialisation manager, who has to act as a conduit

between government and university policy, and the needs of industry and academic

scientist. The process of undertaking this role, however suggested much greater use of

network-based governance for commercialisation managers, particularly with regard

to their relationships with academic scientists.

‘Researchers understand that they’ve got to have a conversation with someone

that’s knowledgeable before they take that particular publication. [However]

clearly you can’t have a commercialisation officer company vetting every

publication.’ AMB

The university context also adds another dimension that affects the relationship

between the scientist and commercialisation manager, because the commercialisation

process is an additional role to the one traditionally carried out by academics.

‘Remember as business manager, commercialisation manager, you’ve got no

power to make people do anything, particularly in universities. I mean in

business you say “Okay, if you don’t want to do that, you’ll have to leave the

91/101

organisation.” In the universities “If you don’t want do that, I can’t stop

you.” There’s no power within the system.’ JC

This highlights the varying motives and outcomes that the commercialisation manager

has to manage: the scientist and their motive to publish, the university and their

motive to dissuade the scientist to publish if it will reduce the value of the IP and

potential returns for the university, and the motives of the commercialisation manager

who wants to manage the information to give industry and government a compelling

proposition to attract significant funding. Developing a cooperative culture between

the business units, universities, and government was therefore recognised as integral

to the relationships needed:

‘So you need people in universities – and there are not many of these people –

who can bridge that gap between the science and the science culture and the

business and the business culture.’ PR

In contrast to the hierarchical governance push on universities and industry from

government, commercialisation can occur through spinouts as a way of disseminating

knowledge seemed to be more towards a market-based approach than hierarchical. An

examination of the commercialisation structures utilised by universities themselves,

also highlights a focus on more market-based governance modes, separated from

other parts of university management, with continuing conflicts between this and

more traditional university approaches.

‘Commercialisation is not a core business for the University. That’s why – I

think that’s why [a University name omitted] puts it out into [its

commercialization unit name omitted], because it’s not actually a core

business. Whereas their core business is education, teaching.’ JC

In terms of the impacts on learning processes, however, the inherent challenge for the

academic scientist however, is how to manage the additional commercial

responsibility produced by these changes:

92/101

‘There’s always this tension. When you’re outside the system you think “why

don’t they do this? Surely they want to commercialise what they’ve done”.

But when you’re in a university you’re fighting for grants, you’ve got PhD

students, you’ve got your teaching work, you’ve got your research, you’ve got

your administrative duties, and then they want you to commercialise. Yeah,

and you’ve got insecurity of tenure. …’ AMB

This also had a knock-on effect on perceptions of the structures and fora for

dissemination in place. Specifically respondents recognised similar results in start-up

behaviour in universities, questioning the “forum” (i.e. starts- ups), as a result being

utilised for knowledge dissemination (and utilisation) :

‘And part of [the university] their charter was to create X number of spin-outs

in a certain amount of time…Whether any of them would be useful or not is

another thing and also the state government at that stage had a particular

funding scheme that allowed, [name omitted], to set up spin off companies and

get funding for them. …They have a certain amount of intellectual property;

they’re managed by the head of the [name omitted, commercialisation unit]

plus a commercial development officer who’s working on it part time and it’s

pretty much a cart without wheels. It just sits there and does nothing.’ RIB

The appropriateness of the allocation of funds to public institutions including

universities for research and developments versus private institutions and research

centres was therefore questioned by several industry respondents:

‘You look at the major recipients of funding out of government. Go back and

look at the last seven years, since they announced in 1999 that they were

going to concentrate on biotech. Have a look at all the funding for life

scientists you'll find about 95 percent has gone to universities.’ KA

Following this agenda, government funding and start-up programs have reflected in

changes of start-up behaviour. Several respondents argued that this may just be an

indication of being able to access funds, rather than the policy creating sustainable

firms or growth.

93/101

‘For instance, the state government in grants that invest in commercialisation

of new technologies have to be granted to a company. So you see universities

doing things like forming a small, really, shell company, so they can take an

$80,000 ISIS grant or a $100,000 COMET grant because they have to, not

because it’s necessarily the right vehicle to put that piece of IP in at that point

in time.’ AMB

The evidence presented therefore highlights a number of different knowledge

governance modes at work simultaneously in this industry in knowledge

dissemination processes. Specifically, a strong government hierarchical mode is

seemingly in place between the main stakeholders of government, industry and

universities. Simultaneously there appears to be an incomplete move towards a more

market-based regime for universities in dealing with commercialisation, and

seemingly more reliance on network-based modes within universities between the

scientists and the commercialisation manager, to deal with the additional emphasis for

universities to commercialise their knowledge.

Table 4 : Knowledge Utilisation Processes: Case Study Results

Process Case One (TT) Case Two(PB) Case Three (GT) Knowledge Asymmetry

No No not for management team, Yes between scientist and board

Yes Various forms

Knowledge Relationship EO: Risk Taking for Scientist

Calculated (Remained in position, though uncertain outcomes, support by stakeholders)

Very High Risk (Customer/market driven) Market Leader

Very High (Scientist left job, International set up)

Knowledge Relationship EO: Proactiveness

Very High (Prioritised)

Very High (Left job, mortgaged house)

High (Opportunities recognised, difficulty with timing and funding)

Knowledge Relationship EO: Innovativeness

Suite of Potential Applications for Platform

Very High (Customer/market driven) Market Leader

Product Definition failure

Knowledge Management: Trust

Yes High between founding members

None (Board and CEO)

Knowledge Management: Communication

High Weekly, Informal

High, Weekly, Informal at commencement, Reduced as progress (Board vs Management) More Structured as more established

Low ties with Board (proximity?) High level of communication between staff and management

Knowledge Management:

High High commitment at commencement

High at commencement Low at cessation

94/101

Knowledge Utilisation Processes

The case study firms exhibited different outcomes, product definition, age, and thus

provide a snapshot of differing categories of knowledge asymmetry, knowledge

relationships seen through entrepreneurial orientation and knowledge management.

This did allow, however, analysis of how the factors highlighted as of importance to

knowledge utilisation worked differently in the 3 case study examples, as table 4

illustrates

In comparing the three cases in terms of the key constructs of knowledge asymmetry,

entrepreneurial orientation and knowledge management, the following key differences

can be noted.:-

• The failed case (3) displayed more knowledge asymmetry, compared with

other two. In terms of knowledge relationships, the failed case exhibits, in

particular, a product definition failure in innovativeness, compared with the

other cases. In terms of knowledge management, trust and communication

were particularly poor in the failed company compared with the others.

Knowledge asymmetric behaviour was exhibited in two cases, both between

the scientist and the board of directors with restricted information flow on

strategy and science. This lack of communication has further implications for

accurate strategic decision making and the ‘best way forward’ with the

potential to make less effective decisions as a team.

• All three cases show high levels of knowledge relationships through

entrepreneurial orientation. This is not unusual based on the inherent nature of

the biotechnology industry. High levels of pro-activeness and risk taking

behaviours were seen in cases two and three, with the first case shown

moderate risk behaviour.

• There are interrelationships not only between knowledge management

characteristics of trust, communication and commitment but also between

elements of communication and commitment and asymmetric behaviours, risk

and trust constructs, and knowledge asymmetry and perceived risks.

Commitment

95/101

To summarise, therefore, unlike cases one and two, case study three experienced

faults particularly in knowledge asymmetry, knowledge relationships and knowledge

management characteristics. The three cases also highlight the complexity of the

interrelationships between the factors in these biotechnology firm start-ups.

Proximity Issues

In terms of the role of local and cross-location national and international linkages with

a variety of stakeholders for knowledge creation, there are specific issues related to

attracting and retaining talent in the Australian biotechnology industry highlighted by

Stephens et al (2006). Fontes (2006) has also recently highlighted that biotechnology

firms also form collaborator relations with ‘distant networks’ (Fontes, 2006) to

augment access to their own research (often-non networked) knowledge spillovers

from their own localities.

Knowledge Dissemination Overall

In terms of knowledge dissemination, the interviews with key industry stakeholders

highlighted that the role and importance of geographical clustering of the industry was

not clear-cut:-

‘What is happening is that with the [name omitted] and the [name omitted]

and a few other smaller features we are getting clustering but it’s not in the

like industry. So you are going to get a few biotechnology firms coming

together and we are still getting minor cross fertilisation and some synergy

but not to the extent of the actual clustering theory.’ JK

This also highlights the use of universities in knowledge spillover indirectly through

the provision of centralised facilities, education and training. This may assist in

innovation dissemination in ways other than spinouts, university proximity also being

discussed in terms of providing (agglomerational) access to resources.

‘You are also finding more satellite-like clusters coming out of universities. I

don't think it’s an issue of dependency on the universities. It’s more like a

security blanket, of the university is right there, and from the scientists who

utilise not only the human capital but also the equipment capital.’ JK

96/101

Knowledge Dissemination Cases

Finally, in terms of knowledge dissemination, table 5 below indicates that for cases

one and two international linkages seemed to be beneficial to utilization. For case

three, however, the cross-local linkages seemed to increase knowledge asymmetric

behaviours, re-emphasising the lack of effective knowledge management in this case.

Table 5: Proximity Issues in the Case Studies Title Case One (TT) Case Two(PB) Case Three (GT) PROXIMITY ISSUES

International Linkages

Use of International and bonding networks (informal) in product development, bridging networks in business development

During start up, limited international links as business developed further extensive links including international markets and further links including international partners in continuing product development

International bonding for processes for product development, and sought international markets. Potential to move internationally with pressure from the board increased asymmetric behaviours within the team and created further complications during business development.

Thus, the role and importance of proximity seems to differ depending on whether the

process in question is knowledge creation, dissemination, or utilization based. In

addition, for utilization processes, it may also be interdependent with knowledge

asymmetry, EO, and knowledge management processes at work within individual

companies.

CONCLUSIONS

These results indicate the heavily government-influenced nature of Australian

biotechnology industry-policy both in knowledge creation and knowledge

dissemination mechanisms, the evidence suggesting a very hierarchy-based approach

from government. This contrasts with the more market-based approach towards which

university management of spinout processes seems to be moving, and the seemingly

more network governance based approach that commercialisation managers are

employing in their dealings with the academic scientists, highlighting hybrid

governance currently at work, with respondents questioning the effectiveness of such

behaviour. Commercialisation managers are the conduit of information flows between

97/101

government and policy, the central university research program and academic

scientist. They therefore need to separately manage a plethora of complex

relationships being generated from disparate motivations and stakeholders seeking

differing outcomes, using and being affected by different governance modes. These

include the relationship between the commercialisation unit and scientist, the

relationship between the commercialisation unit and the central university research

programs. In terms of knowledge utilisation processes, the case-study evidence

indicates that the factors identified in the literature were of both of relevance and are

interlinked both with each other and also with issues related to the impact of local and

cross-locational linkages.

Broadly, the data presented indicates that the factors identified from the literature in

terms of knowledge creation, dissemination and utilisation are of relevance. There is,

however, also a clear need for further research which allows a more developed

understanding of how the factors interact with each other and with issues of

proximity. In particular, a wider range of firm cases needs to be examined in order to

examine more fully the constructs of knowledge asymmetry, entrepreneurial

orientation and knowledge management, in order to both analyse their importance and

their interrelationships with knowledge creation and dissemination mechanisms.

In terms of policy for the Australian biotechnology industry, however, the results as

they stand suggest the need for improved dialogue between the stakeholders in such

new firm development, both in the external environment and internal to the firm itself.

Recognising differing research agendas, expectations and motivations of knowledge

creating actors involved, and their perceptions of knowledge dissemination processes,

may also create better appreciation and understanding of knowledge utilisation

outcomes. The role of the university commercialisation manager seems one key area

for focus, as is, potentially, knowledge management within the firm itself.

REFERENCES

Abramson, N. H., Encarnacao, J., Reid, P. P., & Schmoch, U. (1997). Technology Transfer Systems in the United States and Germany. Washington, DC: National Academy Press.

Acs, Z.J., Audretsch, D.B., Braunerhjelm, P. and Carlsson, B. (2004). The Missing Link: The Knowledge Filter and Entrepreneurship in Endogenous Growth, Discussion Paper, No.4783, December, London, UK, Centre for Economic Policy Research.

98/101

Acs, Z., O’Gorman, C., Szerb, L., Terjesen, S. (2007 forthcoming) “Could the Irish miracle be repeated in Hungary”, Small Business Economics, 28 (2-3), 123-142.

Ancori, B., Bureth, A., & Cohendet, P. (2000). The economics of knowledge: the debate about codification and tacit knowledge. Industrial and Corporate Change, 9(2), 255-287.

Asheim, B., and Coenen, T. (2006). Contextualising Regional innovation Systems in a Globalising Learning Economy: On Knowledge Bases and Institutional Frameworks, Journal of Technology Transfer, 31, 163-173.

Asheim, B., and Gertler, M. (2005). The Geography of Innovation: Regional Innovation Systems, in: J Fagerberg, D C. Mowery & R R Nelson (eds), The Oxford Handbook of Innovation, Oxford: Oxford University Press, pp. 291-317.

Audretsch, D. (2004). Sustaining Innovation and Growth: Public Policy Support for Entrepreneurship, Industry and Innovation, 11(3), 167-191.

Audretsch D. and E. Lehmann (2005). Does the Knowledge Spillover Theory of Entrepreneurship hold for regions?, Research Policy, 34, 1191-1202.

Australia, B. (2006). An Australian Government Inititiative: Biotechnology Australia. Ausbiotech (2007). About Biotechnology. Industry Overview.

http://www.ausbiotech.org/content.asp?pageid=25 Baptista, R. and Swann, P. (1998). Do Firms in Clusters Innovate More?, Research Policy,

27(5), 527-540. Baranson, J. (1990). Transnational strategic alliances: why, what, where and how.

Multinational Business, 2, 54-61. Barney, J. (1991). Firm Resources and Sustained Competitive Advantage. Journal of

Management, 17(1), 99-120. Barringer, B. R., & Bluedorn, A. C. (1999). The relationship between corporate

entrepreneurship and strategic management. Strategic Management Journal, 20.

Boschma, R.A. (2005). Proximity and Innovation: A critical assessment, Regional Studies, 39 (1), 61-74.

Braczyk, H.J. and Heidenreich, M. (1998). Conclusion, in H.-J. Braczyk & P. Cooke & M. Heidenreich (eds.), Regional Innovation Systems; The Role of Governances in a Globalized World. London: UCL Press.

Brooksbank, D.J., Jones-Evans, D., Kwong, C.C.Y, Thompson, P., and Ullah, F., (2007). Liquidity constraints upon start-ups with new products:a study of reasons for failing to access finance using Global Entrepreneurship Monitor (2005) Data, National Entrepreneurship Observatory of Wales, Pontypridd. http://64.233.179.104/scholar?hl=en&lr=&q=cache:2mD7IFKvXUcJ:www.utwente.nl/nikos/htsf/papers/ullahone.pdf+(2005)+entrepreneur+Bootstrapping+knowledge+asymmetry

Brusco, S. (1982). ‘The Emilian Model: Productive Decentralization and Social Integration’, Cambridge Journal of Economics, 6, 167-184.

Brusco, S. and Righi, E. (1989). ‘Local Government, Industrial Policy and Social Consensus: the Case of Modena (Italy)’, Economy and Society, 18(4), 405-424.

Carroll, G.R. and Hannan, M.T. (2000). The Demography of Corporations and Industries, Princeton University Press, Princeton.

Cimon, Y. (2004). Knowledge-related asymmetries in strategic alliances. Journal of Knowledge Management, 8(3), 17-30.

Cohen, W. M., and Levinthal, D. A. (1990). Absorptive Capacity: A New Perspective On Learning And Inno. Administrative Science Quarterly, 35(1), 128.

Cooke, P. and Morgan, K. (1994). The regional innovation system of Baden Württemberg, International Journal of Technology Management, 9: 394-429.

Cooke, P., Uranga, M.G., and Etxebarria, G. (1997). Regional innovation systems: Institutional and organisational dimensions, Research Policy, 26: 475-491.

99/101

Cooke, P. (1998). Introduction: Origins of the concept. In H.J. Braczyk, P. Cooke & M. Universitydenreich (Eds.), Regional Innovation Systems – The Role of Governance in a Globalized World. London: UCL.

Cooke, P. Clifton N. and Oleaga, M. (2005). Social capital, firm embeddedness and regional development, Regional Studies, 39(8), 1065–1077.

Cooke, P and De Laurentis, C. (2006). in Cooke, P., De Laurentis, C., Tödtling, F., Trippl, M. Regional Knowledge Economies: Markets, Innovation and Clusters, (Edward Elgar).

Covin, J. G., and Slevin, D. P. (1988). The influence of organization structure on the utility of an entrepreneurial top management style. Journal of Management Studies, 25(3), 217-234.

Covin, J. G., and Slevin, D. P. (1989). The strategic management of small firms in hostile and benign environments. Strategic Management Journal, 10, 75-87.

Davidsson, P. and Delmar, F. (2001). Les entreprises à forte croissance et leur contribution à l'emploi: le cas de la Suède 1987-1996. Revue Internationale PME, 14(3-4), 164-187.

Delmar, F. and Davidsson, P. (1998). ‘A Taxonomy of High-growth Firms’. Paper presented at Babson College-Kaufmann Foundation Entrepreneurship Research Conference. Ghent, Belgium, May 21-23.

Dyer, J. H., Kale, P., and Singh, H. (2001). How to make strategic alliances work. Sloan Management Review, 42(4), 37-43.

Erskinomics Consulting, P. L. (2003). Critical Factors In Successful R&D An International Comparison: A Discussion Paper prepared for The Australian Innovation Association and The Australian Institute for Commercialisation.

Etzkowtiz, H. and Leydesdorff, L. (1997). Universities in the Global Economy: A Triple

Helix of Government-Industry and Government Relations. London, Croom Helm.

Feldman, M.P. (1994). The Geography of Innovation (Kluwer, Dordrecht). Felsentsein, D. And Portnov, B.A. (2006). The liability of smallness: Can We Expect Less

Regional Disparities in Small Countries? In Regional Disparities in Small Countries (Springer, Berlin).

Freel, M.S. (2003). External innovation and Small Firm Performance, Entrepreneurship and Regional Development, 12 (3) 245-266.

Frenz, M., Michie, J. and Oughton, C. (2005). Innovation and cooperation: the role of absorptive capacity, Birkbeck Working Paper Series: Birkbeck University of London

Frenz, M., and Oughton, M. (2006). Innovation in the UK Regions and Devolved Administrations: A Review of the Literature: Final Report for the Department of Trade and Industry and the Office of the Deputy Prime Minister, DTI, London.

Fukuyama, F. (1995) Trust: the social virtues and the creation of prosperity (The Free Press, New York).

Gordon, I.R. and McCann, P. (2000). Industrial Clusters: Complexes, Agglomeration and /or Social Networks, Urban Studies, 37 (3), 513-532.

Goffin, K. and Mitchell, R. (2005). Innovation Management: Strategy and Implementation Using the Pentathlon Framework, (Palgrave Macmillan Ltd, Houndmills).

Harman, G., and Harman, K. (2004). Governments and Universities as the Main Drivers of Enhanced Australian University Research Commercialisation Capability. Journal of Higher Education Policy and Management, 26(2), 153-169.

Havnes, P.A., and Senneseth, K.(2001). A Panel Study of Firm Growth among SMEs in Networks, Journal Small Business Economics, 16 (4), 293-302.

Howells, J. (1999). Regional Systems of Innovation, in: D Archibugi, J Howells and J Michie (eds), Innovation Policy in a Global Economy, Cambridge: Cambridge University Press.

100/101

Howells, J. (2002). Tacit Knowledge, Innovation and Economic Geography, Urban Studies, 39(5-6): 871-884.

Jaffe, A.B., Trajtenberg, M., and Henderson, R. (1993). Geographic localization and knowledge spillovers as evidenced by patent citation, Quarterly Journal of Economics, 108, 577-598.

Kautz, K., and Thaysen, K. (2001). Knowledge, learning and IT Support in a Small Software Company, Journal of Knowledge Management, 5 (4), 349-357.

Keast, R. and Brown, K. (2002). The Government Service Delivery Network: A Case Study of the Push and Pull of Central Government Coordination, Public Management Review, 4(4), 439 – 459.

Kirby, D.A. (2004). “Entrepreneurship education: can business schools meet the challenge?”, Education and Training, 46 (8/9), 510-519.

Lambert, R. (2003). Lambert Review of Business-University Collaboration: Final Report, HM Treasury, HMSO, London.

Lee, C., Lee, K., and Pennings, J. M. (2001). Internal capabilities, external networks, and performance: A study on technology-based ventures. Strategic Management Journal, 22(6/7), 615.

Leydesdorff, L. (2000). The Triple Helix: An Evolutionary Model of Innovation, Research Policy, 29 (2), 243-255.

Liyanage, S., & Barnard, R. (2003). Valuing of Firms' Prior Knowledge: A Measure of Knowledge Distance. Knowledge and Process Management, 10(2), 85-98.

Lowndes, V., and Skelcher, C. (1998). “The dynamics of multi-organisational partnerships: An analysis of changing modes of governance.” Public Administration, 76(2), 313-333.

Lumpkin, G. T., and Dess, G. G. (1996). Clarifying the entrepreneurial orientation construct and linking it to performance. Academy of Management Review, 21(1), 135-172.

Lundvall, B-Å (ed) (1992). National Systems of Innovation: Towards a Theory of Innovation and Interactive Learning (Pinter, London and New York).

Malerba, F. (2002). Sectoral systems of innovation and production, Research Policy, 31, 247-264.

Matlay, H. (2000). Organisational learning in small learning organisations: an empirical overview. Education and Training, 42,(4/5), 202-211.

McCracken, G. (1988). The long interview-Sage University Paper-Qualitative Research Methods Series 13, Beverley Hills : Sage

Menard, C. (2002). The Economics of Hybrid Organizations (Presidential Address ISNIE 2002, MIT September 29, Available at www.isnie.org).

Miller, D. (1983). The correlates of entrepreneurship in three types of firms. Management Science, 29, 770-791.

Morgan, K. (1997). The learning region: institutions, innovation ad regional renewal, Regional Studies, 31, 491-504.

Narula, R. (2003). Globalization and Technology: Interdependence, Innovation Systems and Industrial Policy, (Polity Press, Cambridge, UK).

Nonaka, I., and Takeuchi, H. (1995). The Knowledge Creating Company. (Oxford University Press, New York).

O’Gorman, C. (2000). Strategy and The Small Firm, In: Evans, D.J. and Carter, S. (Eds.), Enterprise and The Small Business: Principles, Practice and Policy, (Pearson Education Ltd, Essex).

Orsenigo, L. (1989). The emergence of biotechnology-Institutions, markets and industrial innovation. (Printer, London).

Oughton, C., Landabaso, M., and Morgan, K. (2002). The Regional Innovation Paradox: Innovation Policy and Industrial Policy, Journal of Technology Transfer, 27(2), 97-110.

Patton, M. Q. (1987). How to Use Qualitative Methods in Evaluation. (Sage Publications, Newbury Park).

101/101

Patton, M. Q. (1990). Qualitative Evaluation and Research Methods. (Sage Publications, Newbury Park).

Pickernell, D., Packham, G., Thomas, B., and Keast, R. (2008). University Challenge? Innovation Policy and SMEs in Welsh Economic Development Policy: Towards a New Framework International Journal of Entrepreneurship and Innovation, 9 (1), 51-62.

Poon, J.M., Ainuddin, R.A., and Junit, S.H. (2006). Effects of self-concept traits and Entrepreneurial Orientation on Firm Performance, International Small Business Journal, 24 (1) 61-82.

Putnam R. (2000) Bowling Alone: The Collapse and Revival of American Community (Simon and Schuster, New York).

Rhodes, R.A.W. (1997). From marketisation to diplomacy: It’s the mix that matters, Public Policy and Administration, 12, 31-50.

Salojarvi, S., Furu, P. and Sveiby, K-E. (2005). Knowledge Management and growth in Finnish SMEs, Journal of Knowledge Management, 9 (2), 103-122.

Schaeffer, P. and Loveridge, S. (2002). Towards and understanding of types of public private cooperation, Public Performance and Management Review, 26 (2), 169-189.

Senyard, J. (2007). Interrelationships between knowledge asymmetry, entrepreneurial orientation and relationships marketing characteristics and impacts on relationship development for Australian Biotechnology firm start ups, unpublished thesis, Queensland University of Technology, Brisbane, Australia.

Simmie, J., Sennet, J., Wood, P., and Hart, D. (2002). Innovation in Europe: A Tale of Networks, Knowledge and Trade in Five Cities, Regional Studies, 36(1), 47-64.

Voss, Z.G., Voss, G.B., and Moreman, C. (2005). An empirical examination of the complex relationships between entrepreneurial orientation and stakeholder support, European Journal of Marketing, 39 (9/10), 1132-1150.

Weick, K.E. (1990). Organizational culture as a source of high reliability. Calif. Manage. Rev. 29 (2), 112–127

Welsh, B. (1996). Developing Managers for the Smaller Business: A Report on Training Development, (Institute of Management, London).

Wickert, A., and Herschel, R. (2001). Knowledge management issues for smaller businesses, Journal of Knowledge management, 5 (4), 329-337.

Wiklund, J., and Shepherd, D. (2005). Entrepreneurial Orientation and small business performance: a configurational approach, Journal of Business Venturing, 20 (1), 71-91.

Wynarczyk, P. and Watson, R. (2005). Firm Growth and Supply Chain Partnerships: An Empirical Analysis of U.K. SME Subcontractors, Journal Small Business Economics, 24 (1), 39-51.

Yeh-Yun-Lin, C. (1998). Success Factors Of Small and Medium Sized Enterprises In Taiwan: An Analysis Of Cases, Journal of Small Business Management, 36(4), 43-56.

Zahra, S. A. (1993). A conceptual model of entrepreneurship as firm behavior: A critique and extension. Entrepreneurship Theory & Practice, 5-21.

Zhou, K.Z., Yim, C.K. and Tse, D.K. (2005). The effects of strategic orientations on technology: and market based breakthrough innovations, Journal of Marketing, 69: 42-60.